# Aterio's Knowledge Base


About Aterio		aboutus.png	ABOUT US
Data Products	Solutions & Datasets	data_product_icon.png	DATA PRODUCTS
Data Delivery Options	Cloud, API & MCP	integration.png	Data Delivery Options
FAQs		faq.png	General FAQs
Glossary		glossary.png	Glossary
Contact Us		support.png	https://www.aterio.io/contact-us

# About Aterio Aterio provides forward-looking data on population, housing, energy, and industrial developments across the U.S. Our datasets empower investors to anticipate demand shifts, identify emerging markets, and capitalize on long-term infrastructure and energy trends. Key datasets include: **New Development Projects Data** * ***Data Centers:***\ Comprehensive dataset tracking active, planned, and proposed data centers, including power capacity, square footage, and development stages. * ***Power Generation Projects:***\ Detailed coverage of new generation, storage, and transmission projects across all ISOs and utilities, with status, capacity, and developer info. * ***New Industrial Developments:***\ Tracks large-scale industrial projects—including logistics, manufacturing, and semiconductor facilities—by location, size, and timeline. * ***Early Signals on New Developments:***\ Live feed of early-stage development activity from land sales, zoning applications, and utility filings to corporate announcements. This data empowers AI agents to run automated workflows that enhance business decision-making. **Location Intelligence Data** * ***US Population Forecast:***\ Dynamic, forward-looking population estimates by age and gender cohort at the ZIP code level through 2035 in a dynamic respond. * ***US Housing Forecast:***\ Projected housing unit growth at the ZIP code level, aligned with population. Our data is updated hourly or daily and is available via **Snowflake, S3, GCP, Databricks**, or as downloadable **CSV files**—making it easy to integrate, share, and analyze. Additionally, for certain products, our data is available via **API and MCP**. Discover more about these products in our **Data Products** section. *** *Designed for speed and precision, Aterio’s data powers real-time alerts for investors and executives, and enables AI agents to trigger workflows and notify decision-makers as new developments occur.* # RESOURCES # About the Aterio Investment Score What is the Aterio Score? The Aterio Score is a comprehensive tool used to evaluate real estate investment opportunities by assigning a numerical value to various influencing factors. This score is generated from six key indexes: Prices, Population and Migration, Demand vs. Supply, Economic Development, Capacity to Pay, and Risk. By aggregating these factors, the Aterio Score provides a structured and data-driven approach to real estate decision-making, helping investors objectively compare different locations and make informed choices. The Prices Index examines real estate price fluctuations and market dynamics. The Population and Migration Index assesses the impact of migration patterns on local markets. The Demand vs. Supply Index evaluates the balance between property availability and market demand. The Economic Development Index considers local economic growth and its effects on real estate. The Capacity to Pay Index measures the financial capability of potential buyers and renters. Finally, the Risk Index evaluates market volatility and other potential risks. This scoring system enables investors to identify high-potential neighborhoods, understand market trends, and manage risks effectively, ultimately supporting strategic investment decisions ([Aterio](https://www.aterio.io/articles/what_is_the_aterio_score)) ([Aterio](https://www.aterio.io/articles?category=Score+and+Indexes)) ([Aterio](https://www.aterio.io/real-estate)). # About our Local Risk Index What is the Local Risk Index by aterio? The Location Risk Index is an integral part of the Aterio Score, designed to assess various risk factors that could impact real estate investments in a specific area. This index provides a numerical score that reflects the potential risks associated with investing in a particular location. It is calculated by analyzing a range of critical indicators, including unemployment rates, foreclosure rates, the availability of clean drinking water, and local news about plant closures and layoffs. Understanding unemployment rates helps gauge economic stability in the area, as high unemployment can lead to decreased demand for housing. Foreclosure rates indicate financial distress among homeowners, which can affect property values. The availability of clean drinking water is a crucial factor for livability and can impact long-term property desirability. Additionally, news about plant closures and layoffs provides insights into local economic conditions and potential future risks. By incorporating these factors, the Location Risk Index offers a comprehensive view of the potential challenges and pitfalls in a given real estate market. A high-risk score might signal a need for caution, whereas a lower risk score could indicate a more stable investment environment. This index enables investors to make more informed decisions by understanding and managing the risks associated with different locations, ultimately supporting strategic real estate investment planning # Data Centers "US Data Centers" product provides a comprehensive inventory of data centers across the United States, including those currently operational, under construction, and announced for future construction. Sourced from data center providers, press releases, satellite imagery, energy grid data, Aterio’s proprietary energy curves, and utility company records. This dataset offers detailed information on locations, sizes, utility companies serving those markets, and estimated power demand. Some of the tickers covered by the dataset include LUMN, DLR, EQIX, GOOG, NTTYY, RXT, META, AMZN, AAPL, MSFT, ORCL, and others. To provide the most comprehensive and detailed information on each Data Center building and campus, we have established the following data tables:


Inventory	data-centers-inventory.png	Inventory
Events	data-centers-events.png	Events
Estimated Energy Consumption	data-centers-consumption.png	Estimated Power Consumption
Balancing Authority Energy Demand	data-centers-demand.png	Balancing Authorities Energy Demand

# Datasets # Inventory This data is part of our comprehensive Data Centers Dataset, offering customers an in-depth look at data center properties, including facility sizes, provider details, and power capacity estimates. Perfect for stakeholders seeking insights into the evolving data center industry, it provides all the necessary details to assess infrastructure at both individual and aggregated levels. ### Schema Overview

Field Name	Type	Description
ATERIO_DATA_CENTER_UID (PK)	String	Unique identifier for each data center, used to track and reference facilities across various datasets.
DATA_CENTER_BUILDING_NAME	String	Name of the building where the data center is located.
ATERIO_DATA_CENTER_CAMPUS_UID	String	Unique identifier for each data center campus, used to track and reference facilities across various datasets.
DATA_CENTER_CAMPUS_NAME	String	Name of the campus or complex where the data center is situated, if applicable.
DATA_CENTER_STAGE	String	Project stage (announcement, construction, active, cancelled, delayed, land bank), representing the current lifecycle phase.
LATEST_SATELLITE_PICTURE_DATE	Date	The most recent date when a satellite image of the data center was captured, useful for tracking construction progress or operational status.
PCT_CONSTRUCTION_STATUS	Float	The estimated percentage of construction completed, providing insight into the progress of the data center project.
FLG_AI_FACILITY	String	Indicates if there is public data indicating the data center will be an AI facility: Possible values: Y/N/Null. Null does not imply the facility is non-AI. It simply means we abstain from classifying unless there is direct language in filings, press releases, or regulatory documents specifying AI/HPC use.
FLG_BTM_ONSITE_POWER_GENERATION	String	Indicates if there is public data indicating the project will support behind-the-meter (BTM) power generation. Possible values: Y/N/null
DATA_CENTER_ACTIVATION_DATE	Date	The official or estimated date when the data center became fully operational and available for use.
ESTIMATED_ACTIVE_DATE_BY	String	The source or method used to estimate the expected activation date of the data center (e.g., developer projection, industry analysis).
ATERIO_PROVIDER_UID	String	Unique identifier for the data center provider in our systems.
PROVIDER_NAME	String	Name of the company or entity that owns or operates the data center.
PROVIDER_URL	String	URL to the provider's main website.
PROVIDER_PUBLIC_PRIVATE	String	Status indicating whether the provider is public or privately owned.
PROVIDER_TICKER_NAME	String	Stock ticker symbol for the provider.
PROVIDER_BLOOMBERG_TICKER_NAME	String	Bloomberg ticker symbol for the provider, for financial data tracking.
STOCK_EXCHANGE_PROVIDER_NAME	String	Stock Ticker including Stock Exchange
PROVIDER_BACKED_BY	String	Corporations related to Provider / Investors backing the Provider
CONSTRUCTION_EQUIPMENT_PROVIDER_COMPANIES	String	Suppliers of machinery and equipment for data center construction
PROJECT_FINANCING_COMPANIES	String	Financial institutions providing capital and funding for the project
END_USER_COMPANIES	String	Organizations that will occupy, lease, or utilize the completed data center
TOT_PROJECT_COST	Integer	Total cost for the project.
FULL_ADDRESS	String	Complete address of the data center.
ZIP_CODE	String	Postal code for the data centers location.
COUNTY_FIPS_CODE	String	Federal Information Processing Standards (FIPS) code identifying the county.
COUNTY_NAME	String	Name of the county where the data center is located.
CITY_NAME	String	City where the data center resides.,
PLACE_FIPS_CODE	String	FIPS code for the specific place (city/town).
STATE_CODE	String	Abbreviated state code (e.g., CA for California).
STATE_NAME	String	State Name
COUNTRY_CODE	String	Abbreviate country code (e.g., US for United States, CA for Canada)
COUNTRY_NAME	String	Country Name
LOCATION_LATITUDE	Float	Latitude coordinate of the data center's location.
LOCATION_LONGITUDE	Float	Longitude coordinate of the data center's location.
SITE_ACREAGE	Float	Total land area occupied by the data center, measured in acres.
TOT_FACILITY_SPACE_SQFT	Integer	Total facility size in square feet.
TOT_DATACENTER_SPACE_SQFT	Integer	Square footage specifically dedicated to the data center space.
PROV_PUB_TOT_POWER_CAPACITY_MW	Float	Total published power capacity available for the data center, measured in megawatts (MW).
ATERIO_EST_TOT_POWER_CAPACITY_MW	Float	Estimate of total power capacity by Aterio (MW).
ATERIO_EST_TOT_POWER_CAPACITY_MW_LOWER	Float	Lower estimate of total power capacity by Aterio (MW).
ATERIO_EST_TOT_POWER_CAPACITY_MW_UPPER	Float	Upper estimate of total power capacity by Aterio (MW).
SELECTED_POWER_CAPACITY_MW	Float	Selected total power capacity depending on the availability of both published and estimated values.
AVG_MARKET_POWER_COST	Float	Average power market cost
YEARLY_PUE	Float	The yearly average Power Usage Effectiveness (PUE) of the data center, measuring its energy efficiency (lower values indicate higher efficiency).
TOT_NUM_GENERATORS	Integer	The total number of backup power generators installed at the data center, indicating its redundancy and resilience in case of power failures.
ATERIO_ELECTRICAL_SUBSTATION_UID	String	Unique identifier for the electrical substation serving the data center.
SUBSTATION_NAME	String	Name of the electrical substation.
SUBSTATION_FULL_ADDRESS	String	Full address of the substation.
SUBSTATION_ZIP_CODE	String	ZIP code of the substation.
SUBSTATION_COUNTY_FIPS_CODE	String	FIPS code identifying the county of the substation.
SUBSTATION_COUNTY_NAME	String	Name of the county where the substation is located.
SUBSTATION_CITY_NAME	String	City where the substation resides.
SUBSTATION_PLACE_FIPS_CODE	String	FIPS code for the specific place (city/town).
SUBSTATION_STATE_CODE	String	State code of the substation.
SUBSTATION_LOCATION_LATITUDE	Float	Latitude of the substation's geographic location.
SUBSTATION_LOCATION_LONGITUDE	Float	Longitude of the substation's geographic location.
ATERIO_ELECTRICAL_UTILITY_UID	String	Unique identifier for the utility company supplying electricity.
UTILITY_NAME	String	Name of the utility company.
UTILITY_CODE	String	Code assigned to the utility for identification purposes.
UTILITY_PUBLIC_PRIVATE	String	Status indicating whether the utility company is public or privately owned.
UTILITY_TICKER_NAME	String	Stock ticker symbol for the utility company, if publicly traded.
UTILITY_BLOOMBERG_TICKER_NAME	String	Bloomberg ticker symbol for the utility provider, for financial data tracking.
UTILITY_EXCHANGE_PROVIDER_TICKER_NAME	String	Ticker name for the stock exchange where the utility company is listed.
ATERIO_BAL_AUTH_UID	String	Unique identifier for the balancing authority responsible for the transmission and distribution of electricity to the data center.
BAL_AUTH_ABBR	String	Abbreviated name or acronym of the balancing authority.
BAL_AUTH_NAME	String	Full name of the balancing authority responsible for managing the electricity grid in the region where the data center is located.
ATERIO_BAL_AUTH_SUBREGION_UID	String	Unique identifier for the balancing authority responsible for the transmission and distribution of electricity to the data center.
BAL_AUTH_SUBREGION_CODE	String	Code of the subregion under the balancing authority’s management.
BAL_AUTH_SUBREGION_NAME	String	Name of the subregion under the balancing authority’s management.
DATASHEET_URL	String	URL linking to a datasheet with additional information.
MAP_URL	String	URL linking to an interactive map showing the data center location.
PROJECT_PERMIT_URL	String	URL linking to the project permit information.
CAPEX_URL	String	URL linking to documentation detailing the project's capital expenditure (CAPEX).
PROJECT_EXECUTION_LIKELIHOOD	String	A categorical indicator of the probability that a data center project will be executed, based on available information such as provider announcements, market conditions, and project stage. Possible values: High, Medium, Low.
NOTES	String	Free-form field for additional comments or remarks about the data center.
RECORD_CREATED_DATE	Date	The date when the record was first created.
RECORD_UPDATED_DATE	Date	The date when the record was last modified or updated.
UPDATED_AT	Date	The date when the last execution or process update occurred.

# Events This data offers a historical view of notable events in the data center industry, providing insights into key developments for data centers and their providers. Whether it's announcements, expansions, or operational updates, the Data Center Events table keeps customers informed on important milestones. ### Schema Overview

Field Name	Type	Description
ATERIO_DATA_CENTER_UID (PK)	String	Unique identifier for each data center, used to track and reference facilities across various datasets.
DATA_CENTER_CAMPUS_NAME	String	Name of the campus or complex where the data center is situated, if applicable.
DATA_CENTER_BUILDING_NAME	String	Name of the building where the data center is located.
ATERIO_PROVIDER_UID	String	Unique identifier for the data center provider in our systems.
PROVIDER_NAME	String	Name of the company or entity that owns or operates the data center.
EVENT_TYPE (PK)	String	Event type. Possible values: "Active", "Construction Start", "Construction End", "Announcement", "Inactive"
EVENT_DATE	Date	Event date.
UPDATED_AT	Date	The date when the last execution or process update occurred.

# Estimated Power Consumption This table tracks estimated power consumption in the data center industry, providing insights into energy usage trends and capacity changes. ### Schema Overview

Field	Type	Description
ATERIO_DATA_CENTER_UID (PK)	String	Unique identifier for each data center, used to track and reference facilities across various datasets.
DATA_CENTER_CAMPUS_NAME	String	Name of the campus or complex where the data center is situated, if applicable.
DATA_CENTER_BUILDING_NAME	String	Name of the building where the data center is located.
ATERIO_PROVIDER_UID	String	Unique identifier for the data center provider in our systems.
PROVIDER_NAME	String	Name of the company or entity that owns or operates the data center.
ATERIO_ELECTRICAL_SUBSTATION_UID	String	Unique identifier for each electrical substation.
SUBSTATION_NAME	String	Name of the electrical substation
ATERIO_ELECTRICAL_UTILITY_UID	String	Unique identifier for each electrical utility company.
UTILITY_NAME	String	Name of the electrical utility provider.
UTILITY_CODE	String	Code representing the electrical utility.
DATE (PK)	Date	First day of the month, representing the reporting period.
MONTH	String	Month corresponding to the reporting period (format: YYYY-MM)
TOT_POWER_CAPACITY_MW	Float	Total power capacity of the data center (MW).
TOT_SUBSTATION_LOAD_MW	Float	Total load on the substation (MW).
TOT_DATA_CENTER_ESTIMATED_CONSUMPTION_MW	Float	Estimated power consumption of data centers (MW).
UPDATED_AT	Date	The date when the last execution or process update occurred.

# Balancing Authorities Energy Demand This dataset provides critical insights related to balancing authorities and balancing authority subregions. It offers the energy demand in a daily basis for ERCOT, CAISO, MISO, SPP, ISO-NE and PJM. ### Schema Overview

Field Name	Type	Description
ATERIO_BAL_AUTH_UID	String	Unique identifier for the balancing authority responsible for the transmission and distribution of electricity to the data center.
BAL_AUTH_ABBR	String	Abbreviated name or acronym of the balancing authority.
BAL_AUTH_NAME	String	Full name of the balancing authority responsible for managing the electricity grid in the region where the data center is located.
ATERIO_BAL_AUTH_SUBREGION_UID (PK)	String	Unique identifier for the subregion under the balancing authority’s jurisdiction.
BAL_AUTH_SUBREGION_CODE	String	Name of the subregion under the balancing authority’s management.
BAL_AUTH_SUBREGION_NAME	String	Name of the subregion under the balancing authority’s management.
DATE (PK)	Date	Date on which the power consumption data was recorded.
TOT_ENERGY_DEMAND_MWH	Float	Total energy demand by the balancing authority subregion based on the hourly demand in UTC time.
UPDATED_AT	Date	The date when the last execution or process update occurred.

# Data Access Our Data Centers data product is available on the following platforms:


AWS	S3 Bucket	data-centers-access-aws.png
Snowflake Data Cloud	Data Listing	data-access-snowflakepng.png
Aterio's Download Portal	Customers Only	data-access-intenral.png
Databricks		databricks-logo-access.png
GCP	Cloud Storage & BigQuery	gcp-logo-access.png

[Contact us](https://www.aterio.io/contact-us) to schedule a call and learn more about access options and pricing details. # Our Modeling Approach # Data Collection Aterio captures development activity through a combination of public and observable sources: * Press releases and company websites: * Satellite imagery to verify real-world progress * Utility and interconnection filings * Permitting data from local and federal agencies * Manual validation by analysts before anything is published Only structured, non-sensitive information is included. No scraping, no AI-generated data, and no private sources. ### How We Monitor Development: Verifying Buildout Progress and Power Deployment Aterio uses high-resolution satellite imagery to monitor the construction progress of hyperscale data centers across the U.S. Our team analyzes each image to identify structural milestones and infrastructure readiness, including generator installation and substation completion. Based on public infrastructure filings, Microsoft commonly deploys Caterpillar C175-16 generators across its U.S. data centers. A clear example is the campus at 5707 Loving Trail in Cheyenne, WY, which follows the company’s standardized hyperscale layout. Key features include: * 20 diesel backup generators, each rated at 3MW, positioned on dedicated concrete pads * A fully built utility substation for grid interconnection and on-site power distribution * A modular construction layout designed for phased expansion. #### **Visual Progress Example** Aterio tracks changes over time using historical satellite snapshots: One facility has its shell completed and generator installation is halfway done; the rest of the site shows minimal pad preparation.

Image dated May 23, 2023 captured via Google Earth

Two additional structures are almost completed with 20 fully installed generators. A third structure is midway through steel framing, and a fourth has completed foundations.

Image dated March 29, 2025 captured via Pléiades Neo satellite, © 2025 Airbus DS, sourced through SkyWatch

These updates are not generated by AI, but manually verified by analysts following strict visual benchmarking. Example construction note: {% hint style="success" %} \[2025-05-14 JR] Construction: Steel framing is underway with partial roof installation on the western section of the structure. {% endhint %} This method ensures every record in our dataset is tagged with verified construction status and includes timestamps that can be used for forecasting capacity growth, estimating activation dates, and tracking operator expansion pace. # Project Stage Definitions A guide to how Aterio classifies data center projects across their development lifecycle. ***Announcement*** Project has been publicly announced, filed, permitted, or otherwise identified, but no construction activity has been confirmed. ***Construction*** Construction activity has been confirmed through satellite imagery, permits, filings, or public disclosures. This can include early-stage site clearing, grading, foundations, vertical construction, or equipment installation. ***Land Bank*** Land has been acquired or controlled for potential future data center development, but there are no confirmed near-term construction plans. ***Not Approved / Withdrawn*** Project has not received required approvals or has been withdrawn from the approval process by the developer. ***Cancelled*** Project has been formally cancelled by the developer or is no longer expected to move forward, even though initial groundwork or construction progressed. ***Delayed*** Project remains active but has experienced material delays relative to expected development timelines, based on construction progress, public disclosures, or other available evidence. ***Active*** Facility is operational or in final commissioning stages with critical infrastructure installed and capacity expected to be available for use. # Construction Stages & Equipment Installation ### Our Approach Aterio tracks every U.S. and Canadian data center project at the individual building level using satellite imagery, permit filings, county records, and local media. Each building is assigned a construction completion percentage based on visual benchmarks verified by our analysts. This document focuses on **when and how equipment installation becomes** visible during the construction lifecycle - the stages most relevant to estimating activation timelines. #### Site Preparation (5%-15%) * 5% - Site clearing - vegetation and existing structures removed * 10% - Land grading and leveling in progress * 15% - Foundation work underway (concrete footings and pads) #### Structural Construction (20%-40%) * 20% - Vertical construction begins - initial steel framing rising * 30% - Structural steel framing completed * 40% - Roof installation begins #### Exterior & Mechanical Fit-Out (50%-60%) * 50% - Roof installation completed * 60% - Roof fully installed; mechanical equipment being placed #### Equipment Installation (70%-95%) * 70% - External building prep begins (equipment pads, trenching, facade work) * 85% - Generator pads populated; substation equipment connected * 90% - Semis visible at loading bays - IT hardware being moved in * 95% - Full mechanical and electrical commissioning underway {% hint style="warning" icon="circle-info" %} **KEY INSIGHT** Above 70%, external equipment becomes visible via satellite: cooling towers, generators, and transformer yards. At 85–90%, semi-trailers at loading docks signal IT hardware delivery. {% endhint %} ### Equipment Installation Timeline *What becomes visible at each stage as observed via satellite imagery.* {% stepper %} {% step %} **External Equipment Appears (70%)** Cooling towers, chillers, and backup generator units begin appearing on dedicated concrete pads adjacent to and on top of the building. Transformer yards and switchgear enclosures become visible near the substation interconnect. {% endstep %} {% step %} **Generator Installation Ramps Up (85%)** Diesel backup generators (typically 2–3 MW units such as Cat C175-16) are installed on pre-cast pads. Generator count becomes countable from satellite, enabling Aterio’s ML model to estimate facility power capacity. {% endstep %} {% step %} **IT Hardware Delivery Begins (90%)** Semi-trailers appear at loading bays, indicating servers, networking gear, and rack infrastructure are being moved into the building. This is a strong signal that the facility is approaching commissioning and initial power-on. {% endstep %} {% step %} **Final Commissioning (95%)** Building structure and all core equipment fully installed. Final site work, testing, and punch-list items underway. Facility is near-ready for activation. {% endstep %} {% endstepper %} ### Early & Mid Construction **Stages: 15%-40%** Multiple buildings at varying stages of early construction. The top structure shows partial roof installation over completed steel framing (\~40%). The middle building has foundations and early vertical steel (\~20%). The bottom structure shows foundation pads and footings being poured (\~15%). No external equipment is visible at these stages.

*Figure: AWS – Project Rainier. Captured via Pléiades Neo satellite, © 2025 Airbus DS*

### Equipment Installation Begins **Stages: 70%-85%** Roof fully completed on both buildings. Initial cooling towers are being installed on the rooftop. The generator yard between the structures is staged with concrete pads but no generators have arrived yet. Medium voltage transformer systems are installed, preparing the electrical infrastructure for generator connection in the next phase.

### Fully Equipped & Near-Activation **Stages: 90%-95%** Buildings are fully equipped. Dense arrays of cooling units cover the rooftop and perimeter. All backup generators are installed and countable. Semi-trailers are visible at loading bays, indicating IT hardware (servers, networking equipment, racks) is being moved into the facility. The site is paved with completed access roads - a strong signal the facility is approaching activation.

# Power Capacity Estimation ML Model ## **Introduction** Data centers are critical infrastructure that house powerful computing resources used for storing, processing, and managing vast amounts of data. The power capacity of a data center refers to the total amount of electrical power it can consume to operate its IT equipment and supporting infrastructure, such as cooling systems and power conditioning units. This capacity is a crucial factor in determining a data center's ability to handle workloads and maintain uptime. The US Data Center Power Capacity Estimation Model is an approach designed to accurately predict the power capacity of data centers across the United States. This model leverages a dataset that includes detailed information on facility size, IT infrastructure, and others. By employing an ensemble modeling technique, we integrate multiple predictive algorithms to enhance the accuracy and reliability of our estimates. This model serves as a vital tool for data center operators, energy planners, and policymakers, providing them with the insights needed to optimize energy usage and plan for future capacity needs effectively. ## **About Data Center's Power Capacity** The power consumption of data centers is significantly influenced by various factors, with artificial intelligence (AI) being a major contributor to recent increases. According to the latest sustainability report from Meta, 14,975 GWh has been consumed by their data centers by the end of 2023. For instance, the data center located in Prineville, OR consumed 1,243 GWh and by applying a simple formula we can estimate the power capacity of the facility. Power Capacity (MW) = Energy Consumption (MWh)Time (hours) x Percentage Utilization Assuming the data center operates continuously over the year (8,760 hours) and utilizes a percentage utilization rate between 70% and 90%, we can estimate a power capacity between 157 MW and 200 MW. Furthermore, as projected by Newmark, power usage will reach 35 GW by 2030, nearly double the 17 GW in 2022. This growth highlights the critical need for precise power capacity estimates to support advanced technologies. Many existing facilities face challenges adapting to new energy and cooling demands, leading to a surge in new construction aimed at meeting these higher performance standards, which directly influences how companies estimate power requirements to stay competitive. ## **Our Power Capacity Estimation Methodology** This section outlines the methodology employed to estimate the power capacity of data centers across multiple providers and locations, considering various project stages, including announced, active, and under construction. On our last recent update, we have developed 2 estimation models: one for hyperscalers and one for colocation, enterprise data centers (non-hyperscalers). ### Power Capacity Estimation Model for Non-Hyperscalers #### **Data Compilation** We compiled a comprehensive dataset that includes critical variables affecting power capacity, such as facility size, power consumption, provider details, geo proximity between data centers and operational status. This dataset serves as the foundation for our analysis. #### **Outlier Detection** Recognizing the significance of accurate data representation, the initial phase of the model focused on identifying potential outliers. Outliers can be defined as data points that exhibit unexpected behavior, deviating from the expected linear relationship between facility size and power capacity. These anomalies may arise from advancements in cooling technologies or server efficiencies, leading to facilities consuming less power relative to their size. With this, we aimed to enhance the robustness of the modeling process.

Insights were gained from identifying potential outliers among data centers that are currently under construction, those that have been announced, and the latest active facilities. These buildings deviate from the expected linear correlation between facility size and power capacity, likely due to the new innovations these companies are implementing to enhance efficiency. #### **Model Evaluation** Following the outlier detection, we employed a series of regression models to evaluate their performance in predicting power capacity. The primary metrics for model performance were the coefficient of determination (R²) and the Mean Squared Error (MSE), which quantifies the proportion of variance in the dependent variable (power capacity) that can be explained by the independent variables (facility size and other factors). The MSE provides a measure of how much the predictions deviate from the actual values, indicating the model's accuracy. An ensemble learning model achieving the highest performance, with an R² value of 0.97 and an RMSE of 12.72, was selected for further analysis. **Prediction Generation** Utilizing the selected model, we generated predictions for the power capacity of the remaining data centers. To enhance the interpretability of these predictions, we also established a prediction interval with a confidence level of 80%. This interval provides a range within which we expect the true power capacity of each data center to lie, accounting for the inherent uncertainty in our estimates. #### **Insights** The following chart illustrates the relationship between the data center size and the power capacity, including our new estimations:

Power Capacity vs Data Center Space (SqFt)

In data centers that are under construction or in the announcement phase, the lines are steeper, suggesting higher power requirements as the facility size increases. This aligns with expectations as newer data centers may have greater demands for power infrastructure due to more extensive and modern equipment being installed. For those in the active phase, the slope is more moderate, reflecting potentially more optimized and stable power usage patterns. This differentiation makes sense when considering the evolving nature of data centers. Recent advancements in technology, particularly in artificial intelligence (AI) chips and the more efficient distribution of IT workloads, have contributed to an overall trend of achieving more computational power without the need for dramatically increasing physical space. These innovations, combined with power management improvements, are creating more power-efficient data centers, allowing for greater productivity and performance in less space, which is reflected in the tilting trend lines. ### Power Capacity Estimation Model for Hyperscalers This second model focuses exclusively on data centers identified as *Hyperscaler —* large-scale, efficient facilitities designed to support cloud and computing services for companies like Amazon or Google. As of July 2, 2025, we have identified the 1,514 buildings, with 47% currently Active:

#### **Data Compilation** Key variables for this model are: * Number of Generators * Project Stage * Facility Size * US State * Provider * Data Center campus vs Data Center building The number of generators are estimated using satellite pictures (Review *Data Collection* section for more details). #### **Outlier Detection** On this scenario, we applied the same tecniques to remove any potential outlier record, minimizing the risk of overfitting or underfitting. #### **Model Evaluation** We applied the same evaluation metrics (R2 and RMSE) for this estimation model, which quantifies the proportion of variance in the dependent variable (power capacity) that can be explained by the independent variables including the number of generators. The final ensemble model achieved an R² of 0.94 and an RMSE of 10.0182. **Prediction practices** are consistent with those used in the non-hyperscaler model. **Insights** In previous releases, we used the same estimation model for both non-hyperscalers and hyperscalers. This new model improved accuracy specially for major providers like AWS, Google, Microsoft and Meta. The estimated power capacity has increased to reflect recent announcements and deeper analysis of generator requirements, aligning better with current industry trends.

Comparison between models (for Hyperscalers)

# Expected Power Capacity Demand Projection The rapid expansion of data centers across the United States is driving a surge in electricity demand to power advanced computing, including AI, cloud services, and enterprise workloads. To better anticipate this trend, we use an **expected power capacity methodology**. This methodology goes beyond simply recording announced or active capacities. It provides a **scenario-based perspective** on how much power is likely to come online over time, offering a more realistic view of future demand. Since not every announced project will proceed as planned, we incorporate **project likelihood ratings**. Facilities are grouped into three levels: * **"High" likelihood** – Projects are highly likely to progress to buildout and operation due to firm land control with permitting and zoning nearly complete, secured or committed power infrastructure, strong financial backing, experienced developers or anchor tenants. * **"Medium" likelihood** –Projects have a medium likelihood of progressing to buildout and operation due to lacking land ownership, unconfirmed power infrastructure, early-stage permitting, new or less experienced developers, community concerns, and limited existing data center presence. * **"Low" likelihood** – Projects have a low likelihood of progressing to buildout and operation due to incomplete or disputed land control, stalled or denied permitting, unavailable or uncertain power infrastructure, little to no published financial backing, strong regulatory or community opposition, and developers with little or no track record. ### The Three Scenarios 1. **Optimistic Case** * Assumption: Every data center project comes online as planned. * Purpose: Represents the maximum possible capacity if the industry expands without major obstacles. * Use Case: Useful for high-growth planning and understanding the upper bound of demand.
2. **Baseline Case** * Assumption: Each project is weighted by its likelihood of execution. For example, projects considered “high probability” contribute fully to the estimate, while those with “medium” or “low” probability are discounted. * Purpose: Provides a more realistic picture of expected growth by balancing ambition with execution risk. * Use Case: Serves as the primary projection for investors, policymakers, and planners.
3. **Conservative Case** * Assumption: Only projects with a “high” likelihood of execution are included. * Purpose: Shows the minimum level of new capacity likely to materialize. * Use Case: Helps organizations plan for worst-case scenarios and ensure resilience. ### Cumulative Growth Over Time Capacity is not only assessed on a yearly basis but also tracked cumulatively. This means we can see how states, providers or utility companies build up total capacity over multiple years.

Expected Power Capacity Demand (GW) in Virginia

### Conclusion This methodology provides a structured way to project data center power capacity demand by balancing optimism with realism. By offering three scenario paths—optimistic, baseline, and conservative—we help stakeholders prepare for both rapid growth and more cautious expansion, ensuring that infrastructure, investment, and policy keep pace with one of the fastest-growing industries in the energy and technology landscape.
# Data Quality Aterio follows a structured validation process to ensure that key figures such as total power capacity, number of facilities by state, or provider-specific footprints are accurate and aligned with industry disclosures. Each data point is verified using a combination of public sources, including **utility filings**, **10-Q reports from public companies**, **permitting records**, and **satellite imagery**. For power capacity, we confirm generator counts and models through permitting data and visual inspection. These estimates are then cross-checked against manufacturer specifications and official provider disclosures. For example, if *Microsoft* reports using 3MW Caterpillar C175-16 generators, we validate both the type and number installed to ensure our totals reflect reality. At the regional level, we reconcile our data center counts with regulatory filings, operator announcements, and public infrastructure documents. When a provider claims a specific footprint in a state, we locate and verify each facility individually to ensure that our numbers are both granular and consistent with what the industry reports.
# Use Cases #### **IT Infrastructure Planning** Optimize data center placement, capacity, and timing using Aterio’s detailed project-level data—including development stage, power estimates, and square footage. Site selection and capacity planning are informed by real-time insights into where hyperscale and colocation facilities are expanding across the U.S. **Energy Consumption Analysis** Leverage Aterio’s modeled power demand by site and operator to analyze regional energy needs, load profiles, and sustainability gaps. The dataset supports utilities, grid planners, and ESG teams in understanding the operational footprint of current and future data center clusters. **Market Analysis** Assess competitive density and future supply by analyzing Aterio’s inventory of announced, under-construction, and active data centers. Filter by provider, location, or megawatt scale to identify underserved markets, regional saturation, or strategic partnership opportunities. **Policy Analysis** Support state and federal energy, land use, and infrastructure policy by tracking where data center growth is concentrated and how much power it will require. Aterio’s dataset enables proactive planning around energy resilience, water usage, and emissions associated with digital infrastructure. # Analyzing Construction Activity & Market Momentum This page explains how Aterio builds monthly momentum metrics for US data centers, using the latest inventory and events data. These outputs are used in dashboards and can be replicated or adapted by clients in their own environments. We currently generate 2 main calculations: 1. Under Construction Momentum 2. Announced Momentum #### Source Tables The momentum logic is based on two core datasets: * **Inventory** * `ATERIO_DATA_CENTER_UID (PK)` * `COUNTRY_CODE` * `STATE_CODE` * `SELECTED_POWER_CAPACITY_MW` * `TOT_DATACENTER_SPACE_SQFT` * **Events** * `ATERIO_DATA_CENTER_UID` * `EVENT_TYPE` (e.g. Announced, Construction Started, Active, Cancelled, Not Approved / Withdrawn) * `EVENT_DATE` #### Under Construction Momentum **Goal:** For each month end, estimate how much capacity and square footage is actively under construction. **Logical Rules** For a given facility and a given month `AS_OF_MONTH`, it is considered Under Construction if: * `construction_start_date IS NOT NULL (not empty value)`, and * `construction_start_date <= AS_OF_MONTH`, and * `activation_date IS NULL (empty value) OR activation_date > AS_OF_MONTH` *In other words, construction has started, but the facility is not yet active.* We then aggregate by `STATE_CODE (or any other dimension)` and `AS_OF_MONTH`: * `q_under_construction_facilities` – count of facilities * `s_under_construction_power_mw` – sum of `SELECTED_POWER_CAPACITY_MW` * `s_under_construction_square_footage` – sum of `TOT_DATACENTER_SPACE_SQFT` **Example** This chart illustrates under-construction momentum: bars indicate facility counts, and the line reflects MW capacity.

**SQL Script**

WITH dc_dates AS (
  SELECT
    inv.ATERIO_DATA_CENTER_UID,
    inv.STATE_CODE,
    inv.SELECTED_POWER_CAPACITY_MW AS power_capacity_mw,
    inv.TOT_DATACENTER_SPACE_SQFT  AS square_footage,
    MIN(IF(ev.EVENT_TYPE = 'Construction Started', ev.EVENT_DATE, NULL)) AS construction_start_date,
    MIN(IF(ev.EVENT_TYPE = 'Active',               ev.EVENT_DATE, NULL)) AS activation_date
  FROM `[Data Center Inventory Table]`  inv
  LEFT JOIN `[Data Center Events Table]` ev
    ON ev.ATERIO_DATA_CENTER_UID = inv.ATERIO_DATA_CENTER_UID
  WHERE inv.COUNTRY_CODE = 'US'
  GROUP BY 1,2,3,4
),
month_series AS (
  SELECT
    LAST_DAY(d, MONTH) AS asof_month
  FROM UNNEST(GENERATE_DATE_ARRAY(
        DATE '2021-01-01',
        CURRENT_DATE(),
        INTERVAL 1 MONTH)) AS d
),
under_construction AS (
  SELECT
    ms.asof_month,
    dd.STATE_CODE,
    COUNTIF(
      dd.construction_start_date IS NOT NULL
      AND dd.construction_start_date <= ms.asof_month
      AND (dd.activation_date IS NULL OR dd.activation_date > ms.asof_month)
    ) AS q_under_construction_facilities,
    SUM(
      IF(
        dd.construction_start_date IS NOT NULL
        AND dd.construction_start_date <= ms.asof_month
        AND (dd.activation_date IS NULL OR dd.activation_date > ms.asof_month),
        dd.power_capacity_mw,
        0
      )
    ) AS s_under_construction_power_mw,
    SUM(
      IF(
        dd.construction_start_date IS NOT NULL
        AND dd.construction_start_date <= ms.asof_month
        AND (dd.activation_date IS NULL OR dd.activation_date > ms.asof_month),
        dd.square_footage,
        0
      )
    ) AS s_under_construction_square_footage
  FROM month_series ms
  CROSS JOIN dc_dates dd
  GROUP BY 1,2
)
SELECT
  STATE_CODE,
  asof_month,
  FORMAT_DATE('%Y-%m', asof_month) AS asof_month_yyyy_mm,
  q_under_construction_facilities,
  s_under_construction_power_mw,
  s_under_construction_square_footage
FROM under_construction
WHERE FORMAT_DATE('%Y-%m', asof_month) != FORMAT_DATE('%Y-%m', CURRENT_DATE())
ORDER BY STATE_CODE, asof_month;

#### Announced Momentum **Goal:** For each state and month end, estimate the announced pipeline that has not yet moved into construction, activation, cancellation, or withdrawal. **Logical Rules** For a given facility and `AS_OF_MONTH`, it is included in Announced Momentum if: * `announced_date IS NOT NULL (not empty value) and announced_date <= AS_OF_MONTH`, and * `construction_start_date IS NULL (empty value) OR construction_start_date > AS_OF_MONTH`, and * `activation_date IS NULL (empty value) OR activation_date > AS_OF_MONTH`, and * `cancelled_date IS NULL (empty value) OR cancelled_date > AS_OF_MONTH`, and * `project_withdrawn_date IS NULL (empty value) OR project_withdrawn_date > AS_OF_MONTH` *So we capture the “pure” announced pipeline that is still live and has not yet progressed or been cancelled.* **Example** This chart illustrates announced momentum: the bars indicate facility counts, and the line reflects MW capacity.

**SQL Script** ```sql WITH dc_dates AS ( SELECT inv.ATERIO_DATA_CENTER_UID, inv.STATE_CODE, inv.SELECTED_POWER_CAPACITY_MW AS power_capacity_mw, inv.TOT_DATACENTER_SPACE_SQFT AS square_footage, MIN(IF(ev.EVENT_TYPE = 'Construction Started', ev.EVENT_DATE, NULL)) AS construction_start_date, MIN(IF(ev.EVENT_TYPE = 'Announced', ev.EVENT_DATE, NULL)) AS announced_date, MIN(IF(ev.EVENT_TYPE = 'Active', ev.EVENT_DATE, NULL)) AS activation_date, MIN(IF(ev.EVENT_TYPE = 'Cancelled', ev.EVENT_DATE, NULL)) AS cancelled_date, MIN(IF(ev.EVENT_TYPE = 'Not Approved/Withdrawn', ev.EVENT_DATE, NULL)) AS project_withdrawn_date FROM `[Data Center Inventory Table]` inv LEFT JOIN `[Data Center Events Table]` ev ON ev.ATERIO_DATA_CENTER_UID = inv.ATERIO_DATA_CENTER_UID WHERE inv.COUNTRY_CODE = 'US' GROUP BY 1,2,3,4 ), month_series AS ( SELECT LAST_DAY(d, MONTH) AS asof_month FROM UNNEST(GENERATE_DATE_ARRAY( DATE '2021-01-01', CURRENT_DATE(), INTERVAL 1 MONTH)) AS d ), announcements AS ( SELECT ms.asof_month, dd.STATE_CODE, COUNTIF( dd.announced_date IS NOT NULL AND dd.announced_date <= ms.asof_month AND (dd.construction_start_date IS NULL OR dd.construction_start_date > ms.asof_month) AND (dd.activation_date IS NULL OR dd.activation_date > ms.asof_month) AND (dd.cancelled_date IS NULL OR dd.cancelled_date > ms.asof_month) AND (dd.project_withdrawn_date IS NULL OR dd.project_withdrawn_date > ms.asof_month) ) AS q_announcements, SUM( IF( dd.announced_date IS NOT NULL AND dd.announced_date <= ms.asof_month AND (dd.construction_start_date IS NULL OR dd.construction_start_date > ms.asof_month) AND (dd.activation_date IS NULL OR dd.activation_date > ms.asof_month) AND (dd.cancelled_date IS NULL OR dd.cancelled_date > ms.asof_month) AND (dd.project_withdrawn_date IS NULL OR dd.project_withdrawn_date > ms.asof_month), dd.power_capacity_mw, 0 ) ) AS s_announcements_power_capacity, SUM( IF( dd.announced_date IS NOT NULL AND dd.announced_date <= ms.asof_month AND (dd.construction_start_date IS NULL OR dd.construction_start_date > ms.asof_month) AND (dd.activation_date IS NULL OR dd.activation_date > ms.asof_month) AND (dd.cancelled_date IS NULL OR dd.cancelled_date > ms.asof_month) AND (dd.project_withdrawn_date IS NULL OR dd.project_withdrawn_date > ms.asof_month), dd.square_footage, 0 ) ) AS s_announcements_square_footage FROM month_series ms CROSS JOIN dc_dates dd GROUP BY 1,2 ) SELECT STATE_CODE, asof_month, FORMAT_DATE('%Y-%m', asof_month) AS asof_month_yyyy_mm, q_announcements, s_announcements_power_capacity, s_announcements_square_footage FROM announcements WHERE FORMAT_DATE('%Y-%m', asof_month) != FORMAT_DATE('%Y-%m', CURRENT_DATE()) ORDER BY STATE_CODE, asof_month; ``` #### How Clients Can Use These Metrics Typical use cases include: * Tracking pipeline evolution over time by state or provider * Identifying markets with growing construction momentum vs. stalled projects * Comparing announced vs under-construction ratios to assess project conversion * Building charts of monthly MW in Announced / Under Construction / Active stages*.* # FAQs # What Are Data Centers? And Why Do They Matter in 2025? A data center is a physical facility that houses the hardware and systems necessary to store, process, and distribute massive volumes of data and digital information. From cloud-based platforms to AI models to social media and streaming services, data centers are the physical backbone of the digital world. Without them, the internet–and much of modern life–would simply not function. ### **Characteristics of a Data Center** | Servers | Physical or virtual machines that process data and run applications | | --------------- | ------------------------------------------------------------------------------------------------------ | | Storage Systems | Devices that store digital information securely (e.g., SSDs, HDDs) | | Networking Gear | Routers, switches, and firewalls that connect systems and control traffic | | Cooling Systems | Keep temperatures stable to prevent equipment from overheating | | Power Supply | Backup generators, UPS (uninterruptible power supply), and power distribution units ensure reliability | | Security | Includes physical security (guards, cameras, access controls) and cybersecurity defences | ### **Types of Data Centers** There are four main types of data centers, each with its own purpose and usage.
1. **Cloud Data Centers** Operated by third-party providers such as [AWS](https://aws.amazon.com/) or [Microsoft](https://www.microsoft.com/en-ca/). They offer scalable, on-demand computing resources over the internet. They’re ideal for rapid deployment and global access without the burden of owning infrastructure. However, organizations must trade off some control, potential compliance concerns and cost concerns due to their OPEX-heavy model. 2. **Enterprise Data Centers** Internally and privately owned, and usually on-premise or in a designated facility. These facilities give organizations full control over their infrastructure. They’re preferred in highly regulated industries for their security and customization, but require significant upfront investment and are slower to scale. 3. **Colocation Data Centers** Colocation data centers provide a middle ground, allowing companies to place their hardware in a third-party facility, offering retention of physical control over their systems, physical security and reliable infrastructure without the cost of owning a building. However, onsite maintenance may require travel and coordination. 4. **Edge Data Centers** Smaller and geographically distributed, edge data centers bring computing power closer to the end user to reduce latency, which is crucial for real-time services like IoT. They provide fast, localized processing but have limited capacity and scalability. ### **Why they matter** Data centers are the backbone of modern technology and are indispensable in today’s hyper-connected world. They are behind every digital action, from [Google](http://cloud.google.com) searches to bank transfers, and training generative AI models. As digital systems become more intelligent, automated, and globalized, data centers scale to meet exponentially growing demands. They enable AI and Big Data; Modern AI models require enormous amounts of computational resources and power. Training and deploying them depend on high-density, low-latency data centers with accelerated hardware. Industries like finance, healthcare, and government rely on data centers for privacy, uptime, and disaster recovery. These centers store everything from medical records, bank data, entire government databases, and more. Data centers also need to be built physically and digitally secure; with cyber threats and geopolitical tensions rising, the location and protection of data infrastructure matter more than ever. They do this with cameras, guards, biometrics, firewalls, encryption, and sophisticated threat detection. ### **How Aterio Fits Into The Equation** As IT environments grow increasingly hybrid—blending cloud services with on-prem infrastructure—visibility becomes a serious challenge. That’s where [Aterio](https://www.aterio.io/) comes in. Our continually updated d[ataset](https://www.aterio.io/datasets/lst_us_data_centers) includes detailed records of colocation centers worldwide, enabling teams to identify, track, and optimize their infrastructure landscape. {% hint style="info" %} **References** * Chang, Yi, and Vishal Damojipurapu. “Why Unified Observability Is the Future of Infrastructure Management.” Equinix, 18 June 2025, * Lin, Jon. “How AI is Influencing Data Center Infrastructure Trends in 2025.” Equinix, 8 January 2025, . * “New Research by Rackspace Technology Reveals Hybrid Cloud and AI Integration as Key Drivers for IT Innovation in 2025.” Rackspace Technology, 14 January 2025, [https://www.rackspace.com/en-ca/newsroom/new-research-rackspace-technology-2025-state-cloud-report. Accessed 24 June 2025](https://www.rackspace.com/en-ca/newsroom/new-research-rackspace-technology-2025-state-cloud-report). * Pacheco, Matt. “The Future of Hybrid Cloud Adoption: Expert Insights for 2025.” TierPoint, 14 January 2025, * Subramanian, Krishna. “Why the enterprise data center will thrive in the age of AI and hybrid cloud computing.” Data Center Dynamics, 19 March 2025, . * “2025 Global Data Center Outlook.” JLL, . * “Understanding the Different Types of Data Centers in 2025.” GBC Engineers, 9 April 2025, . * Vincent, Matt. “8 Trends That Will Shape the Data Center Industry In 2025.” Data Center Frontier, 6 January 2025, . {% endhint %} # What is an AI Data Center? What Are AI Data Centers? AI data centers are specialized facilities built to handle the extreme computing demands of artificial intelligence, especially for tasks like training large language models (LLMs), running generative AI, and processing huge datasets in real time. They deploy specialized chips such as GPUs, TPUs, or custom ASICs. They feature high-speed interconnects enabling fast communication between processors. And they consume immense power, often requiring advanced cooling systems like liquid immersion. ### **How They Are different from traditional data centers** | Feature | Traditional Data Centers | AI Data Centers | | ------------ | ------------------------------------- | ------------------------------------------------------ | | Primary Use | Web hosting, email, and cloud storage | AI/ML model training, inference | | Hardware | CPU-based servers | High-density GPU or TPU clusters | | Power Demand | Moderate | Very high (10–50+ MW per facility) | | Cooling | Air cooling standard | Advanced liquid cooling, immersion systems | | Network | General-purpose bandwidth | Low-latency, high-throughput fabric (e.g., Infiniband) | ### **Why they are so important** AI data centers enable models like [ChatGPT](https://openai.com/index/chatgpt/), [Google Gemini](https://gemini.google.com/), [Meta AI](https://www.meta.ai/), etc. to function. These models need thousands of GPUs across weeks or months to train, which are only available in these specific data centers. Furthermore, hyperscalers like [Microsoft](https://www.microsoft.com/en-ca/), [Google](https://cloud.google.com/?pli=1\&authuser=2), and [Amazon](https://aws.amazon.com/) are investing billions to expand AI capabilities and capacity. This means that they’re reshaping Infrastructure investment as governments and companies are racing to accommodate the needs of these data centers. In a world that’s rapidly adopting AI, these data centers are crucial, and without them, model development would stall, AI adoption in healthcare, education, climate modelling, etc., would slow. Sovereignty over digital infrastructure would shift back toward a few dominant tech firms. ### **What’s next?** We can expect continued growth in hyperscale data centers, which are designed to handle massive amounts of data. We can also expect massive investments from companies like Meta and Microsoft. We’ll likely see more regional AI data centers to reduce latency and increase resilience. Furthermore, we’ll likely see a push for carbon-neutral AI, using renewable energy to power the data centers. Expect new regulations around energy reporting, data use, and labour practices. ### **Ethics** Contrary to regular data centers, AI data centers require sophisticated liquid cooling systems which consume enormous amounts of water, with some using up millions of litres a day. Furthermore, AI data centers consume massive amounts of energy and electricity, especially during model training. Estimates suggest that training GPT-3 required 1.3 gigawatt-hours, equivalent to powering 120 U.S. homes for a year. This consumption raises concerns about carbon emissions and climate change. {% hint style="info" %} **References** * Farney, Melissa. “AI, Data Centers, and the Next Big Correction: Will Growth Outpace Market Reality?” Data Center Frontier, 28 February 2025, . * Jonker, Alexandra, and Alice Gomstyn. “What Is an AI Data Center?” IBM, 21 February 2025, . Accessed 26 June 2025. * Lebowitz, Michael. “Fueling AI Data Centers: Behind The Meter Solutions- Part 1.” Real Investment Advice, 25 June 2025, . Accessed 26 June 2025. * Minevich, Mark. “AI Data Centers And The New Era Of Unprecedented Demand.” Forbes, 28 October 2024, . * Yang, John, and Juliet Fuisz. “The growing environmental impact of AI data centers' energy demands.” PBS, 25 May 2025, . Accessed 26 June 2025. {% endhint %} # The Impact of AI on Power Grids What Investors Need to Know ### AI Integration into Grid Operations Artificial intelligence is increasingly woven into the day-to-day operations of modern power grids. Utilities are using AI for short-term demand forecasting and real-time load balancing, analyzing vast datasets (e.g. smart meter readings, weather, and usage patterns) to predict electricity demand and optimally dispatch generation resources. This helps grid operators better integrate variable renewable energy sources by anticipating fluctuations and adjusting power flows accordingly. AI-driven analytics also enable “self-healing” grid capabilities, where algorithms can detect faults or voltage anomalies instantaneously and autonomously reconfigure the network to isolate outages and reroute power, improving reliability. In addition, predictive maintenance has emerged as a critical AI application: machine learning models process sensor data from transformers, cables, and other equipment to forecast failures before they happen. For example, researchers at [Argonne National Laboratory](https://www.anl.gov/) developed AI software that predicts the remaining useful life of grid components, allowing utilities to replace aging parts before they break and cause blackouts. Such predictive maintenance not only averts unplanned outages but also saves money by optimizing maintenance schedules. In short, AI is becoming the “brains” of the smart grid, helping operators forecast demand, balance loads, and maintain equipment more effectively than ever before. ### Surging Electricity Demand from AI and Data Centers While AI offers efficiencies on the operational side, the proliferation of AI itself is driving major growth in electricity consumption. Training and running AI models require massive computational power, which demands huge amounts of electricity. Global power demand from data centers is projected to double by 2030, reaching roughly 945 terawatt-hours (TWh). About as much electricity as the country of Japan consumes today. The[ International Energy Agency](https://www.iea.org/) (IEA) attributes most of this surge to the rapid uptake of AI workloads in data centers, noting that electricity use for AI-focused data centers could quadruple between 2023 and 2030. As a result, data centers are poised to become a dominant force in power demand growth. In the United States, the influence is especially striking: according to aterio's data, data centers are on track to contribute nearly half of the growth in U.S. power demand through 2030. By that year, Americans will consume more electricity for running data centers and processing data than for the entire U.S. manufacturing of iron, steel, cement, and other energy-intensive goods combined. This sharp rise in demand from AI-centric computing has implications for grid planning and reliability. Data centers typically draw power 24/7, and clusters of new server farms can strain local grids that were not designed for such concentrated load growth. The IEA warns that about 20% of planned new data center projects could face delays getting connected to the grid, due to bottlenecks in expanding grid capacity. Indeed, utilities in regions with major data center growth are feeling the pressure. One report cautions that U.S. electricity demand from data centers could more than double in the next decade, potentially threatening grid reliability if infrastructure upgrades lag. Transformers and transmission lines in fast-growing tech hubs may become chokepoints if they cannot be built or expanded fast enough to meet the surge in AI-related load. There is also a climate dimension: if new demand outpaces the build-out of carbon-free generation, it could slow the transition to a cleaner grid by forcing greater use of existing fossil-fueled power plants. On the positive side, data center operators are investing in efficiency (such as advanced cooling systems) and renewable energy procurement to mitigate these impacts. New cooling technologies can significantly reduce a data center’s energy per computation. However, analysts caution that AI demand is likely to outpace such efficiency gains, meaning absolute electricity usage by data centers will keep rising. For investors, this trend underscores opportunities in companies supporting data center power infrastructure and renewable energy development, as well as the importance of monitoring how big tech manages its growing energy footprint. ### Investment Opportunities in Grid Modernization and Energy Tech The convergence of AI and energy is creating significant investment opportunities across several sectors. Power utilities and governments are now funnelling capital into modernizing the electric grid to handle both the influx of new demand from EVs, data centers, electrified industry, etc. and the integration of AI-driven operational tools. Investors should consider the following key areas: * Energy Storage: Grid-scale battery storage has become crucial for balancing supply and demand in an AI-enhanced, renewable-rich grid. Batteries can store excess solar or wind power and release it when demand (including AI/data center load) peaks, improving grid flexibility. Global investment in battery energy storage is booming, exceeding $20 billion in 2022 and expected to hit a record $35 billion in 2023. Deployment of grid batteries is accelerating at \~75% annual growth, and the market could more than double in the next several years. Analysts project that by 2030, the battery storage industry will be worth well over $100 billion globally. Companies involved in battery manufacturing stand to benefit from this momentum. Beyond batteries, other storage solutions like pumped hydro and emerging technologies also present investment avenues as the grid’s need for flexibility grows.
* Grid Modernization & Infrastructure: Decades of underinvestment mean that many grid networks require significant upgrades. There is a broad push to build a “smart grid” that is more automated, reliable, and capable of handling distributed energy resources. In the U.S., for example, the federal government has committed billions through the 2021 Infrastructure Investment and Jobs Act to strengthen grid infrastructure. The[ Department of Energy’s](https://www.energy.gov/) Grid Modernization programs will invest up to $3 billion from 2022–2026 in advanced grid technologies. In late 2024, DOE announced an additional $2 billion for projects to expand grid capacity for rising loads from data centers, electrification, and manufacturing, and to harden networks against extreme weather. These public investments are leveraging private capital as well. A signal that companies supplying grid hardware and engineering services will see growing demand. Utilities themselves are directing more spending to grid upgrades.
* Smart Grid Software and AI Solutions: As power systems go digital, software becomes a key investment area. Utilities are prioritizing tools like distributed energy resource (DER) management platforms, grid sensors, and AI-powered forecasting. Startups and major tech firms alike are offering solutions for predicting energy demand, managing outages, and optimizing real-time power flow. One growing innovation is Virtual Power Plants (VPPs), which use software to coordinate small-scale resources—like EV chargers, home batteries, and rooftop solar—to function like a utility-scale power plant. AI is also boosting the performance of solar and battery assets by helping dispatch power when it's most valuable. Investors should watch for companies building these platforms or offering AI-as-a-service to utilities.
* Edge Computing and Grid-Edge Devices: A particularly cutting-edge opportunity lies in grid-edge computing. This refers to deploying computing power and AI algorithms at the edge of the grid; in devices like smart meters, solar inverters, EV chargers, or grid control units, rather than relying solely on central data centers. Edge AI can enable split-second decision-making on the grid, which is vital as systems become more complex with distributed energy resources. Analysts predict the edge computing market in energy will grow nearly 40% annually between 2025 and 2030, reflecting the demand for processing data locally and instantly across critical infrastructure. Investors have taken notice: A standout example is [Utilidata](https://utilidata.com/), which, with Nvidia, has developed AI-enabled smart meters that detect issues like voltage spikes instantly and self-correct. These systems are being piloted by major utilities like Portland General Electric and Duquesne Light. As 100+ million legacy meters are due for upgrade, companies developing smart, connected grid hardware are well-positioned for growth. In summary, the companies that enable intelligence at the grid edge, whether through specialized chips, software, or integrated devices, stand to benefit from the grid’s next wave of modernization.
From an investor’s perspective, the convergence of AI and grid modernization is creating a virtuous cycle of investment: AI’s energy needs are spurring grid upgrades, and those upgrades in turn open avenues for more AI deployment. Sectors like energy storage, power electronics, grid software, and semiconductor companies producing specialized AI chips for energy use are all poised for growth. Even industries adjacent to the grid stand to gain, for example, data center developers and renewable energy providers are partnering to ensure new AI data centers come with dedicated clean power supplies, driving investment in solar, wind, and battery farms. The important takeaway is that the power system’s AI-driven evolution will require billions in capital spending, and investors who strategically back the enabling technologies can potentially ride this wave of energy and digital transformation. ### Risks and Challenges Ahead Despite the promise of AI-enhanced power grids, investors must also weigh several risks and challenges that could impact the sector: * Grid Strain and Reliability Risks: One major concern is whether grid infrastructure can keep up with surging demand and increasingly complex power flows. Many electric grids, especially in North America and Europe, are aging and were built for steadier, more centralized demand profiles. And the combination of rapidly growing electric loads (from AI data centers, electric vehicles, etc.) and more frequent extreme weather events is pushing these grids to their limits. Grid strain can manifest as overloaded transformers, voltage instability, or capacity shortfalls in fast-growing regions. If upgrades to transmission lines and substations do not occur fast enough, the risk of blackouts or local outages increases, which could hurt industries reliant on constant power. For example, according to aterio's tracking of Data Center announcements in 2025, some utilities have already had to delay new data center connections due to insufficient grid capacity, highlighting a potential bottleneck for growth. Investors should be mindful that companies facing such reliability issues or high upgrade costs might see impacts on their financial performance. On the flip side, this risk underpins the opportunity in grid-hardening investments. Firms providing solutions to relieve grid congestion or improve resiliency will be in demand.
* Regulatory and Policy Uncertainty: The power sector is highly regulated, and the advent of AI in grid operations introduces new regulatory questions that are not yet resolved. Utilities and grid operators often need regulatory approval to deploy new technologies or to recover their costs through rates. Currently, many AI applications lack clear regulatory frameworks. Questions about the transparency, validation, and accountability of AI decisions in grid management remain a hurdle. For instance, if an AI system makes a mistake that contributes to an outage, who is liable? Policy bodies like the U.S. [Federal Energy Regulatory Commission](https://www.ferc.gov/) are only beginning to grapple with such issues. A recent policy memo urged FERC and DOE to establish a task force to develop standards for AI in grid planning and operations, noting that current rules are vague on how AI tools can be incorporated in compliance with reliability and transparency requirements. Until standards and best practices are set, some utilities may be hesitant to fully deploy AI solutions beyond pilot projects. Investors should track regulatory developments closely; supportive policies could accelerate adoption, whereas onerous rules or uncertainty could slow it. Additionally, permitting and planning processes for grid expansions can be lengthy; if policymakers streamline these in response to AI-driven demand, that would be a positive signal for investors in infrastructure.
* Cybersecurity Threats: As power grids digitalize and connect more devices, the attack surface for cyber threats expands dramatically. A grid increasingly guided by AI and software is potentially vulnerable to hackers who might attempt to manipulate data or control signals. The risk is twofold: hackers can use AI themselves to find and exploit weaknesses in grid control systems, and flaws or biases in AI algorithms could be exploited to disrupt operations. Alarmingly, cyberattacks on energy infrastructure have already multiplied in recent years. The IEA reports that attacks on energy utilities tripled in the last four years, with adversaries employing more sophisticated techniques, including AI, to probe defenses. A successful cyber intrusion could cause widespread outages or damage to equipment, posing a material risk to utilities and their customers. This threat is not theoretical; in 2015 and 2016, cyberattacks caused blackouts in Ukraine, and in 2023, a U.S. regional grid operator reported an attempted breach of critical control systems. AI integration can both exacerbate and help mitigate this risk. On one hand, more grid automation means more pathways for malicious code if not properly secured. On the other hand, utilities are turning to AI for bolstering cyber defense. For example, using machine learning to detect anomalous network traffic or equipment behaviour in real time, enabling faster incident response. Utility executives rank cybersecurity enhancement as one of the top reasons to invest in AI. Going forward, robust cybersecurity measures will be a prerequisite for AI-enabled grid tech, and companies that provide grid cybersecurity solutions are likely to see growing demand. Investors should consider how well-positioned a utility or tech provider is in terms of cybersecurity; those who proactively secure their AI systems may avoid costly breaches and gain a competitive edge in trust.
* Other Challenges (Workforce and Technical Risks): Finally, it’s worth noting some additional challenges that span both operational and investment realms. Workforce and expertise gaps are one concern – power engineering teams need data scientists and AI specialists, who are in short supply at many utilities. In surveys, utilities cite a lack of in-house expertise and high integration costs as barriers to deploying AI solutions at scale. This could slow adoption or lead to higher project costs than anticipated. There’s also the risk of unproven technologies: not every AI pilot will translate into a successful, scalable product. Investors should be discerning about hype versus reality in AI energy startups. Technologies like advanced grid AI or new battery chemistries carry execution risk; some may not perform as expected outside controlled trials, or they may face longer adoption timelines due to cautious utility procurement cycles. Data privacy and AI ethics in grid management may emerge as public issues too. These softer risks underscore the importance of diversification and due diligence in any AI-energy investment strategy.\
{% hint style="info" %} **References** * Argonne National Laboratory – “Revolutionizing energy grid maintenance: How AI is transforming the future” (May 28, 2024)[anl.gov](https://www.anl.gov/article/revolutionizing-energy-grid-maintenance-how-artificial-intelligence-is-transforming-the-future#:~:text=Using%20the%20latest%20in%20artificial,parts%20before%20any%20problems%20occur)[anl.gov](https://www.anl.gov/article/revolutionizing-energy-grid-maintenance-how-artificial-intelligence-is-transforming-the-future#:~:text=Argonne%20scientists%20are%20leveraging%20the,problems%20before%20they%20even%20occur) * International Energy Agency – “Energy and AI” Special Report (Summary) (April 10, 2025)[iea.org](https://www.iea.org/news/ai-is-set-to-drive-surging-electricity-demand-from-data-centres-while-offering-the-potential-to-transform-how-the-energy-sector-works#:~:text=between%20energy%20and%20AI,more%20than%20quadruple%20by%202030)[iea.org](https://www.iea.org/news/ai-is-set-to-drive-surging-electricity-demand-from-data-centres-while-offering-the-potential-to-transform-how-the-energy-sector-works#:~:text=In%20the%20United%20States%2C%20power,demand%20in%20many%20of%20them) * Utility Dive – “Data center, AI load growth could threaten grid reliability: Conference Board” (June 20, 2024)[utilitydive.com](https://www.utilitydive.com/news/data-center-ai-load-growth-grid-reliability-conference-board/719380/#:~:text=Dive%20Brief%3A)[utilitydive.com](https://www.utilitydive.com/news/data-center-ai-load-growth-grid-reliability-conference-board/719380/#:~:text=Global%20energy%20consumption%20from%20data,cited%20by%20Heil%20and%20Pollard) * Renewable Energy World – “Will AI help with electrification and load growth? (Itron survey)” (Oct 7, 2024)[renewableenergyworld.com](https://www.renewableenergyworld.com/power-grid/smart-grids/will-ai-help-with-electrification-and-load-growth-most-utilities-seem-to-think-so/#:~:text=Nearly%2090,as%20the%20top%20use%20cases)[renewableenergyworld.com](https://www.renewableenergyworld.com/power-grid/smart-grids/will-ai-help-with-electrification-and-load-growth-most-utilities-seem-to-think-so/#:~:text=,data%20infrastructure%2C%20governance) * Canary Media – “Can AI chips make the grid smarter? Utilidata & Nvidia” (Apr 29, 2025)[canarymedia.com](https://www.canarymedia.com/articles/climatetech-finance/ai-grid-utilidata-nvidia#:~:text=list%20of%20tasks%20and%20to,both%20cheaper%20and%20better%20since%C2%A0then)[canarymedia.com](https://www.canarymedia.com/articles/climatetech-finance/ai-grid-utilidata-nvidia#:~:text=In%202023%2C%20Portland%20General%20Electric,impacts%20of%20home%20EV%20charging) * World Economic Forum – “What is edge AI – and why is it so important for energy delivery?” (Jun 20, 2025)[weforum.org](https://www.weforum.org/stories/2025/06/edge-ai-resilient-infrastructure-energy/#:~:text=Edge%20AI%20puts%20intelligence%20right,analysts%20predict%20%205%20nearly)[weforum.org](https://www.weforum.org/stories/2025/06/edge-ai-resilient-infrastructure-energy/#:~:text=vehicle%20chargers%20%E2%80%93%20and%20keeps,and%202030%20%E2%80%93%20using%20it) * Scientific American/Nature – “Data Centers Will Use Twice as Much Energy by 2030 — Driven by AI” (Apr 10, 2025)[scientificamerican.com](https://www.scientificamerican.com/article/ai-will-drive-doubling-of-data-center-energy-demand-by-2030/#:~:text=communities%20to%20plan%20infrastructure%20and,AI%20deployment)[scientificamerican.com](https://www.scientificamerican.com/article/ai-will-drive-doubling-of-data-center-energy-demand-by-2030/#:~:text=Countries%20are%20building%20power%20plants,being%20connected%20to%20the%20grid) * U.S. Department of Energy – Grid Deployment Office: Smart Grid Grants program (2022–2024)[energy.gov](https://www.energy.gov/gdo/smart-grid-grants#:~:text=Smart%20grid%20technologies%20funded%20and,governmental%20entities%2C%20and%20tribal%20nations)[energy.gov](https://www.energy.gov/gdo/smart-grid-grants#:~:text=Second%20Funding%20Opportunity) * International Energy Agency – “Global Energy Storage” (Tracking Clean Energy Progress) (2023)[iea](https://www.iea.org/energy-system/electricity/grid-scale-storage#:~:text=Global%20investment%20in%20battery%20energy,capacity%20targets%20set%20by%20governments) * Day One Project (FAS) – “Unlocking AI’s Grid Modernization Potential” (June 25, 2025)[fas.org](https://fas.org/publication/unlocking-ai-grid-modernization-potential/#:~:text=Surging%20energy%20demand%20and%20increasingly,need%20to%20modernize%20our%20system)[fas.org](https://fas.org/publication/unlocking-ai-grid-modernization-potential/#:~:text=Regional%20Transmission%20Organizations%20,leave%20uncertainty%20around%20the%20explainability) {% endhint %} # Power Generation Projects Aterio.io’s **US Power Generation Dataset** provides a comprehensive inventory of power generation and supporting infrastructure projects across the United States, including utility-scale and distributed assets that are operational, under development, or in the planning stages. Sourced from regulatory filings, interconnection queues, developer disclosures, satellite imagery, energy market data, and Aterio’s proprietary models, the dataset offers detailed information on generation technology, project location, developers, utilities, interconnection status, generation capacity, storage capacity, and grid integration. The dataset covers solar, wind, battery energy storage, hydrogen, natural gas generation, and other power generation technologies, as well as related transmission infrastructure. It provides visibility into regional generation capacity additions, project development activity, and power investment trends across major U.S. ISOs, RTOs, utilities, and balancing authorities. # Dataset ### Dataset Components Enhanced data quality through integration of regulatory sources, queue data, and proprietary modeling to reflect true market-ready capacity and grid viability. Technology and stage classification to track each project’s progress from pre-permitting through construction and commercial operation. Geospatial and ISO mapping aligning each project to its utility service area, ISO zone (e.g., CAISO, PJM, ERCOT), and local permitting authorities. ### Daily Updates Stay informed ensuring you always have the most current insights in this rapidly evolving sector. ### Schema Overview | Field | Type | Description | | --------------------------------------------------- | ------- | ---------------------------------------------------------------------------------------------------------------------------------------------------- | | ATERIO\_POWER\_GENERATION\_PROJECT\_UID | String | The unique identifier of the project. | | PROJECT\_NAME | String | The most recent commercial or marketed name used by the developer for the project. | | FLG\_BTM\_PROJECT | Boolean | Indicates whether the project is a behind-the-meter installation primarily serving on-site load. | | DEVELOPER\_COMPANIES | String | The parent companies or developers behind the project. | | DEVELOPER\_COMPANIES\_TICKER | String | Stock ticker symbols for the developer companies, if publicly traded. | | CONSTRUCTION\_EQUIPMENT\_PROVIDER\_COMPANIES | String | The EPC contractors and/or major equipment manufacturers involved in the project. | | PROJECT\_FINANCING\_COMPANIES | String | The financial institutions or project finance providers backing the project. | | CUSTOMER\_COMPANIES | String | The offtakers or end customers purchasing the project's energy, capacity, or credits under a signed agreement. | | TOT\_PROJECT\_COST | String | The total estimated or reported capital expenditure for the project in US dollars. | | PROJECT\_FOOTPRINT\_ACREAGE | String | The land area directly occupied by the project's physical infrastructure and equipment in acres. | | SITE\_BOUNDARY\_ACREAGE | String | The total land area enclosed within the legal or administrative site boundary in acres, including buffer zones and undeveloped portions. | | TOT\_CONTRACTED\_POWER\_CAPACITY\_MW | Float | The total capacity in megawatts under a signed offtake agreement, when available. | | AGREEMENT\_URL | String | A URL to a public source discussing the offtake or commercial agreement, when available. | | AGREEMENT\_TYPE | String | The type of offtake or commercial agreement, when available. | | PROJECT\_SOURCE\_URL | String | The primary public-facing URL for the project, typically the developer's project-specific webpage or the most authoritative public source available. | | PROJECT\_LEVEL\_NOTES | String | Contextual project information not captured elsewhere. | | ATERIO\_POWER\_GENERATION\_PLANT\_UID | String | The unique identifier of the plant. | | PLANT\_NAME | String | The name of the plant. | | EIA\_PLANT\_CODES | String | The plant code(s) as listed in EIA-860, when available. | | EIA\_PLANT\_NAMES | String | The plant names(s) as listed in EIA-860, when available. | | EIA\_ENTITY\_IDS | String | The entity ID(s) assigned by the EIA to the owner(s) or operator(s) associated with the plant. | | EIA\_ENTITY\_NAMES | String | The name(s) of the EIA-registered entity(ies) associated with the plant, corresponding to the EIA Entity IDs. | | FULL\_ADDRESS | String | The street address of the site in which the plant is located. | | ZIP\_CODE | Date | The U.S. postal ZIP code of the site in which the plant is located. | | COUNTY\_FIPS\_CODE | Float | The FIPS code of the county in which the plant is located. | | COUNTY\_NAME | String | The name of the county in which the plant is located. | | PLACE\_FIPS\_CODE | String | The FIPS code of the city or town in which the plant is located. | | CITY\_NAME | String | The city or town in which the plant is located, when applicable. | | STATE\_CODE | String | The abbreviation for the U.S. state in which the plant is located. | | LOCATION\_LATITUDE | String | The geographic latitude of the plant location in decimal degrees. | | LOCATION\_LONGITUDE | String | The geographic longitude of the plant location in decimal degrees. | | ATERIO\_PLANT\_TECHNOLOGY\_TYPE | String | The generation or storage technology employed by the plant. | | ATERIO\_PLANT\_ENERGY\_SOURCE | String | The primary fuel or energy resource used by the plant. | | ATERIO\_ELECTRICAL\_UTILITY\_UID | String | The unique identifier of the local electrical utility associated with the plant, when available. | | UTILITY\_NAME | String | The name of the local electric utility associated with the plant, when available. | | UTILITY\_CODE | String | Code assigned to the utility for identification purposes. | | UTILITY\_PUBLIC\_PRIVATE | String | Status indicating whether the utility company is public or privately owned. | | UTILITY\_TICKER\_NAME | String | Stock ticker symbol for the utility company, if publicly traded. | | UTILITY\_BLOOMBERG\_TICKER\_NAME | String | Bloomberg ticker symbol for the utility company, for financial data tracking. | | UTILITY\_EXCHANGE\_PROVIDER\_TICKER\_NAME | String | Ticker name for the stock exchange where the utility company is listed. | | ATERIO\_BAL\_AUTH\_UID | String | The unique identifier of the balancing authority associated with the plant. | | BAL\_AUTH\_ABBR | String | The abbreviation of the balancing authority associated with the plant, when available. | | BAL\_AUTH\_NAME | String | The name of the balancing authority associated with the plant. | | ATERIO\_POWER\_MARKET\_REGION | String | The electricity market region in which the plant is electrically located. | | PLANT\_SOURCE\_URL | String | The URLs for sources used to populate project, plant, and phase-level data. | | PLANT\_LEVEL\_NOTES | String | Contextual plant information not captured elsewhere. | | ATERIO\_POWER\_GENERATION\_PHASE\_UID | String | The unique identifier of the phase. | | PHASE\_NAME | String | The name assigned to the phase, reflecting its sequence. | | PHASE\_STAGE | String | The current development status of the phase. | | EIA\_GENERATOR\_IDS | String | The EIA-assigned generator ID(s) associated with this phase, as listed in EIA-860, when available. | | TOT\_PHASE\_NAMEPLATE\_POWER\_MW | Float | The total nameplate generation of the phase in megawatts, based on the sum of its associated generators. | | TOT\_PHASE\_NAMEPLATE\_POWER\_SUMMER\_MW | Float | The total summer nameplate generation capacity of the phase in megawatts, reflecting rated output under peak summer conditions. | | TOT\_PHASE\_NAMEPLATE\_POWER\_WINTER\_MW | Float | The total winter nameplate generation capacity of the phase in megawatts, reflecting rated output under peak winter conditions. | | TOT\_PHASE\_STORAGE\_DURATION\_HOURS | Integer | The rated discharge duration of the storage system in hours at full capacity, when available. | | TOT\_PHASE\_STORAGE\_CAPACITY\_MWH | Float | The total energy storage capacity of the phase in megawatt-hours, when available. | | PHASE\_FILING\_DATE | Date | The date when the initial permit or regulatory filings for the phase first appeared. | | PHASE\_ANNOUNCED\_DATE | Date | The date when the phase was publicly disclosed by the developer. | | PHASE\_ATERIO\_ESTIMATED\_CONSTRUCTION\_START\_DATE | Date | Aterio's internal estimate of when physical construction activity began or is expected to begin for the phase. | | PHASE\_ATERIO\_ESTIMATED\_COMPLETION\_DATE | Date | Aterio's internal estimate of when construction for the phase is expected to be completed or was completed. | | PHASE\_ATERIO\_ESTIMATED\_ACTIVATION\_DATE | Date | Aterio's internal estimate of when commercial operation began or is expected to begin for the phase. | | PHASE\_ESTIMATED\_ACTIVATION\_DATE | Date | The publicly reported commercial operation date for the phase. | | PHASE\_CANCELLATION\_DATE | Date | The date when the phase was formally cancelled, withdrawn, or denied approval, if applicable. | | PHASE\_WITHDRAWN\_DATE | Date | The date when the phase was denied regulatory approval or formally withdrawn, if applicable. | | LATEST\_SATELLITE\_PICTURE\_DATE | Date | The date when the most recent satellite image was reviewed by Aterio. | | PCT\_CONSTRUCTION\_STATUS | Float | The estimated percentage of physical construction completed for the phase, based on satellite imagery and available sources. | | RECORD\_CREATED\_DATE | Date | The date when the record was first created. | | RECORD\_UPDATED\_DATE | Date | The date when the record was last modified or updated. | | UPDATED\_AT | Date | The date when the last execution or process update occurred. | # Data Access Our Energy Projects data product is available on the following platforms:


AWS	S3 Bucket	data-centers-access-aws.png
Snowflake Data Cloud	Data Listing	data-access-snowflakepng.png
Aterio's Download Portal	Customers Only	data-access-intenral.png
Databricks		databricks-logo-access.png
GCP	Cloud Storage & BigQuery	gcp-logo-access.png

[Contact us](https://www.aterio.io/contact-us) to schedule a call and learn more about access options and pricing details. # Business Applications **Infrastructure Planning** The Energy Projects dataset enables infrastructure planners to identify and evaluate new generation, storage, and transmission capacity aligned with industrial expansion, hyperscale data center growth, and EV corridor development. By tracking project location, interconnection status, and developer activity, users can assess grid readiness in specific regions and pinpoint where new capacity intersects with rising demand, helping optimize site selection and reduce permitting or transmission constraints. **Clean Energy Investment** Investors and asset managers use the dataset to target clean energy projects—solar, wind, battery storage, and hydrogen—that qualify for federal and state incentives under the Inflation Reduction Act (IRA) and other programs. With visibility into developer profiles, project timelines, and interconnection status, the dataset supports ESG-aligned capital allocation, corporate decarbonization strategies, and due diligence for energy transition assets that offer long-term stability and compliance benefits. **Grid Capacity Forecasting** Grid operators, utilities, and planners rely on Energy Projects data to monitor changes in the generation mix across ISOs and RTOs, helping them forecast supply-demand dynamics, congestion risks, and locational marginal pricing impacts. The dataset enables modeling of grid resilience under various renewable integration scenarios by providing detailed insights into project status, capacity, and technology mix at a regional level—critical inputs for integrated resource planning and transmission expansion. **Market Analysis** Strategy teams and analysts use the dataset to understand regional competition, development pipelines, and long-term capacity additions across all major energy technologies. By tracking trends in permitting activity, developer concentration, and technology mix, the data supports competitive benchmarking, market entry assessments, and M\&A evaluation. This broad visibility across energy markets ensures more informed decisions on capital deployment, partnership strategies, and portfolio diversification. # Our Modeling Approach # Construction & Equipment Installation: Turbines and Reciprocating Engines ### Our Approach Aterio tracks every U.S. power generation project at the individual type of technology level using satellite imagery, permit filings, county records, and local media. Each building is assigned a construction completion percentage based on visual benchmarks verified by our analysts. This document focuses on when and how equipment installation becomes visible during the construction lifecycle - the stages most relevant to estimating activation timelines. #### **Announced Stage (0%)** * 0% - No visible construction activity. The phase remains in the “Announced” stage. #### **Site Preparation (5–20%)** * 5% - Land clearing and vegetation removal. The stage moves to “Construction”. * 10% - Land cleared and levelled. Site preparation complete. * 20% - Early construction footprint established and temporary laydown yards and staging areas appear. #### **Concrete Foundations and Civil Work (30–40%)** * 30% - Excavations and underground pipeline work is visible. The plant layout begins to emerge. * 40% - Concrete pads are poured and major foundation works are largely completed. The final plant layout is clearly identifiable. #### **Major Equipment Installation (50–60%)** * 50% - Machinery and initial equipment is arriving on site. Installation is underway. * 60% - Major general equipment is installed. Installation of the plant’s vertical structures begin, as exhaust stacks and gas containers are built. Electric generators (e.g. turbines, reciprocating engines) begin arriving on site. #### **Vertical Structures and Power Systems (70–80%)** * 70% - Exhaust stacks, air inlet houses, and electric generator installations are completed. * 80% - Electrical yard equipment and wiring are installed, together with transformers. #### **Site Completion (90–100%)** * 90% - All of the plant’s essential operating equipment is installed and ready to be energized * 100% - Sound-dampening fencing is installed, if applicable. The plant is ready to operate. The phase moves to the “Active” stage. ### Longhorn Behind-the-Meter (BTM) Longhorn BTM is the power generation project for the constant electricity supply to the data center with the same name. It’s developed for Crusoe. It is located in Abilene, in the Taylor County (TX), and involves Simple-Cycle Natural Gas turbines as technology for the generation process. #### Civil Works and Equipment Installation **Stages: 30% - 60%** From this image, it can be seen different stages at the same time. On the right side, multiple excavations have taken place in the specific areas set for the natural gas generators. That can be attributed mainly to the installation of the natural gas pipeline, which can be directly appreciated. On the left side, the process has a major advancement in which the foundations are already poured and the turbines’ equipment are in the installation process.

Figure: Satellite image of Longhorn Data Center BTM. Source: Google Earth. Imagery © Google, 2025.

#### Major Equipment Installation and Vertical Structures **Stages: 60%-80%** At this stage, most of the turbines’ equipment is already installed or finishing this process since the turbines and exhausts are in place. There’s no heavy installation machinery. In addition, the vertical installations are already installed. Nonetheless, there’s construction equipment around the facilities.

Figure: Satellite image of Longhorn Data Center BTM. Source: Vantor. Imagery © Vantor, 2026.

#### Fully Equipped & Near-Activation **Stages: 90%-100%** At this stage, the project is completely finished. There is no machinery, construction equipment, and temporary facilities near to the project, the roads completely built and set. However, there’s still opportunity to refine the roads construction in the upper side of the project.

#### Site Plans

Figure: Site plan of Longhorn Data Center BTM. Source: Plans © Crusoe Energy Systems, LLC, 2025.

# New Industrial Developments Aterio.io’s "US Industrial Developments Dataset" offers a detailed inventory of industrial real estate projects across the United States — including warehouses, logistics hubs, and manufacturing facilities that are planned, under construction, or recently completed. This dataset is curated from planning board filings, industrial developers, satellite imagery, utility records, and Aterio’s proprietary industrial development models, providing users with a powerful lens into current and future industrial infrastructure. # Datasets ### Dataset Components Enhanced accuracy and predictive insights based on Aterio’s analytical models that capture regional trends in supply chain infrastructure, industrial buildouts, and logistics expansion. Metadata standardization for linking project stages, facility sizes, and utility zones with geospatial and market data ### Daily Updates Stay ahead of industrial trends with daily updates, ensuring you have the most accurate and timely data for strategic planning. # Inventory This data is part of our comprehensive Industrial Developments Dataset, offering customers detailed insights into project properties, including company information, site characteristics, construction progress, and estimated activation dates. Perfect for stakeholders seeking to evaluate industrial infrastructure at both project-specific and aggregated levels. ### Schema Overview

Field Name	Type	Description
ATERIO_INDUSTRIAL_PROJECT_UID (PK)	String	Unique identifier for each industrial project across datasets.
PROJECT_NAME	String	Name of the industrial development project.
COMPANY_NAME	String	Primary company associated with the project.
COMPANY_TICKER_NAME	String	Stock ticker symbol of the main company, if applicable.
INDUSTRY_TYPE	String	Industry category (e.g., manufacturing, logistics).
PROVIDER_BACKED_BY	String	Corporations or investors backing the project.
FACILITY_OWNER	String	Entity that owns the facility.
COMPANIES_INVOLVED	String	Other companies or partners involved.
PROJECT_STAGE	String	Project phase (announced, construction, active, cancelled, withdrawn).
LOCATION_LATITUDE	Float	Latitude coordinate of the project site.
LOCATION_LONGITUDE	Float	Longitude coordinate of the project site.
FULL_ADDRESS	String	Complete street address of the project site.
ZIP_CODE	String	Postal ZIP code of the project location.
COUNTY_NAME	String	County name where the project is located.
COUNTY_FIPS_CODE	String	Federal Information Processing Standards (FIPS) county code.
CITY_NAME	String	City where the project is located.
STATE_CODE	String	Two-letter U.S. state code.
JOB_CREATION_COUNT	Integer	Expected number of jobs created.
PROJECT_COST	Float	Estimated or published total project cost (USD).
SITE_ACREAGE	Float	Land area of the project site (acres).
TOT_FACILITY_SPACE_SQFT	Float	Total facility space in square feet.
PROJECT_ANNOUNCED_DATE	Date	Official project announcement date.
PROJECT_PROJECT_WITHDRAWN_DATE	Date	Date when the project was withdrawn or cancelled.
PROJECT_CONSTRUCTION_START_DATE	Date	Construction start date.
LATEST_SATELLITE_PICTURE_DATE	Date	Most recent satellite image date for monitoring progress.
PCT_CONSTRUCTION_STATUS	Float	Estimated construction completion percentage.
PROJECT_CONSTRUCTION_FINISHED_DATE	Date	Date construction was completed.
PROJECT_ACTIVATION_DATE	Date	Date the facility became operational.
ESTIMATED_ACTIVE_DATE_BY	String	Source or method used to estimate activation date.
PROJECT_LIKELIHOOD_EXECUTION	Float	Likelihood score of project execution (0–1 scale).
DATASHEET_URL	String	URL linking to project datasheet.
PROJECT_PERMIT_URL	String	URL for project permit or documentation.
CAPEX_URL	String	URL referencing capital expenditure details.
COMMUNITY_RESPONSE_URL	String	URL to community response or feedback.
MAPS_URL	String	URL linking to geographic or map visualization.
SOURCE_URL	String	Original data or publication source.
NOTES	String	Free-form notes or remarks.
RECORD_CREATED_DATE	Date	Date when record was created.
RECORD_UPDATED_DATE	Date	Date when record was last updated.
UPDATED_AT	Date	Timestamp of latest update or dataset refresh.

# Events This data is part of our comprehensive Industrial Developments Dataset, tracking key project events such as announcements, construction milestones, completions, and withdrawals. Ideal for stakeholders looking to monitor project lifecycles and understand the timing and impact of industrial developments. ### Schema Overview

Field Name	Type	Description
ATERIO_INDUSTRIAL_PROJECT_UID (PK)	String	Unique identifier linking the event to a specific industrial project.
PROJECT_NAME	String	Name of the industrial project associated with the event.
COMPANY_NAME	String	Company associated with the industrial project event.
EVENT_TYPE (PK)	String	Type of event (e.g., announcement, groundbreaking, completion, update).
EVENT_DATE	Date	Date when the event occurred or was reported.
RECORD_CREATED_DATE	Date	Date when the event record was initially created.
RECORD_UPDATED_DATE	Date	Date when the event record was last modified.
UPDATED_AT	Date	Timestamp of the latest update or process refresh.

# Data Access Our Industrial Developments data product is available on the following platforms:


AWS	S3 Bucket	data-centers-access-aws.png
Snowflake Data Cloud	Data Listing	data-access-snowflakepng.png
Aterio's Download Portal	Customers Only	data-access-intenral.png
Databricks		databricks-logo-access.png
GCP	Cloud Storage & BigQuery	gcp-logo-access.png

[Contact us](https://www.aterio.io/contact-us) to schedule a call and learn more about access options and pricing details. # Our Modeling Approach # Data Collection Aterio captures industrial development activity across the United States through a combination of public and observable sources: * Press releases, corporate filings, and company websites * Satellite and aerial imagery to verify on-the-ground progress * Federal, state, and local permitting data * Utility and infrastructure filings (power, water, transmission, transport)\ Manual validation by analysts before any record is published All data is sourced from open, non-sensitive records. The dataset excludes AI-generated content, scraping activity, and any private or undisclosed information. ### How We Monitor Development: Verifying Construction and Site Progress Aterio uses high-resolution satellite and aerial imagery to monitor construction progress across major industrial projects, ranging from manufacturing plants and battery facilities to large hospitals, logistics hubs, energy infrastructure, and semiconductor fabs. Our analysts review each site image to identify verifiable construction milestones such as site grading, concrete foundations, steel framing, roof completion, and utility infrastructure. Typical verification markers include: * Site preparation and structural activity (earthwork, foundations, framing) * Utility and infrastructure progress (substations, access roads) * Equipment or systems installation (turbines, generators, tanks, or production lines) ### Visual Progress Example Each project record includes timestamped observations to illustrate progress over time. All updates are reviewed and validated by analysts following strict visual confirmation criteria.

Image dated November 16, 2024 captured via Pléiades Neo satellite, © 2025 Airbus DS

**Example construction note:** {% hint style="success" %} \[2024-11-16 DS] Construction: Structural steel framing advancing across central sections; partial roof installation visible on west and south wings. {% endhint %} This verification process ensures that each entry reflects confirmed, consistent, and timestamped development activity, providing our clients with reliable insight into project momentum, construction progress, and regional industrial growth patterns.
# Industrial Developments Construction Stages Methodology Industrial project durations vary widely depending on scale and complexity. For instance, a 50,000 square foot assembly plant may take around a year to complete, while a semiconductor facility exceeding 500,000 square feet can span more than three years. To ensure consistent progress tracking across these different project types, Aterio applies a percentage-based timeline framework that represents proportional completion milestones rather than fixed durations. ### Percentage Framework Construction milestones are assigned to key progress intervals (e.g., 5%, 10%, 20%). These percentages indicate visible, externally verifiable progress observed through satellite imagery. All timelines account for the reality that interior work constitutes a significant portion of total construction time but remains invisible to satellite monitoring. ### Visual Indicators Used for Progress Assessment Aterio’s analysts measure construction progress using consistent, visible cues in imagery, including: * **Surface transitions** – from vegetation to graded soil to finished surfaces * **Structural evolution** – framing, roofing, and building footprint expansion * **Equipment and exterior installation** – generators, rooftop units * **Site completion work** – paving, landscaping, or boundary development ### Aterio’s Percentages Framework for Industrial Development **5% - Site Clearing Initiated** * Visible: Color change from green/existing to brown/disturbed soil * Indicators: Site boundaries visible, vegetation removed, earth disturbance patterns **10% - Land Grading and Leveling** * Visible: Uniform earth tone, grading patterns, equipment tracks * Indicators: Smooth surface, visible grading contours, drainage features **20% - Foundation Excavation Visible** * Visible: Dark trenches/holes, stockpiles at perimeter * Indicators: Excavation shadows, foundation perimeter defined, soil pile formations **30% - Foundation Concrete Poured** * Visible: Light gray concrete patterns, rectangular foundation footprint * Indicators: Concrete surface texture, defined building outline, uniform gray surface **40% - Structural Steel Erection Begins** * Visible: Shadow-casting vertical elements, steel framework visible * Indicators: Vertical columns creating shadows, grid pattern emerging, height development **50% - Structural Steel Framing Complete** * Visible: Building height evident, structural grid clear * Indicators: Multiple column rows visible, significant shadow casting, building form evident **60% - Roof Structure Begins** * Visible: Mixed roof colors (completed vs in-progress sections) * Indicators: Partial roof coverage, color/texture variations, installation pattern visible **70% - Roof Installation Complete** * Visible: Uniform roof color/texture across entire footprint * Indicators: Complete roof coverage, consistent surface appearance, no open areas **80% - Building Envelope Complete** * Visible: No open bays, uniform building appearance, complete enclosure * Indicators: All exterior walls finished, doors/windows installed, weather-tight structure **90% - Exterior Equipment Pads Visible; Rooftop HVAC Units** * Visible: Rooftop equipment visible, exterior pads/platforms present * Indicators: HVAC units on roof, transformer pads, exterior mechanical equipment **95% - Site Work and Paving Complete** * Visible: Gray/black paved surfaces, parking stripes, loading areas * Indicators: Parking lots finished, access roads paved, line striping visible **100% - Landscaping and Final Site Improvements** * Visible: Clean site, no staging areas/construction equipment, green spaces, defined pathways * Indicators: No visible construction activity, walkways completed, perimeter finished ### Estimated Activation Dates Aterio records the activation or completion date provided directly by the project developer or operator. When such information is not disclosed, activation dates are estimated using standardized construction duration benchmarks to maintain consistency across the dataset. These benchmarks align with typical timelines for large-scale industrial projects and are applied to ensure comparability between industry sectors. #### **Manufacturing Facilities Construction Timeline** Manufacturing facilities represent complex industrial builds due to specialized equipment requirements, utility needs, and process-specific infrastructure. **Timeline by Size** * **Small Facility (<50,000 sq ft):** 10–14 months total construction * **Medium Facility (50,000 – 150,000 sq ft):** 18–24 months total construction * **Large Facility (150,000+ sq ft):** 36–48 months total construction #### **Warehouse Construction Timeline** Warehouses typically have faster construction timelines due to simpler building systems and standardized designs. **Timeline by Size** * **Small Facility (<50,000 sq ft):** 6–9 months total construction * **Medium Facility (50,000 – 150,000 sq ft):** 12–16 months total construction * **Large Facility (150,000+ sq ft):** 22–30 months total construction #### **Solar Farm Construction Timeline** Solar farms have unique construction characteristics with significant equipment installation phases and weather dependencies. **Timeline by Size** * **Small Facility (<10 MW, <100 acres):** 7–10 months total construction * **Medium Facility (10–50 MW, 100–500 acres):** 18–24 months total construction * **Large Facility (50+ MW, 500+ acres):** 30–42 months total construction #### **Energy Projects (Power Plants) Construction Timeline** Power plant construction involves extensive mechanical systems, long-lead equipment, and rigorous safety/environmental requirements. **Timeline by Size** * **Small Facility (<250 MW):** 30–48 months total construction * **Medium Facility (250–800 MW):** 50–72 months total construction * **Large Facility (800+ MW, incl. nuclear):** 70–96+ months total construction #### **Hospitals Construction Timeline** Hospital construction is among the most complex due to stringent codes, specialized systems, and operational requirements during construction. **Timeline by Size** * **Small Facility (<100 beds, <100,000 sq ft):** 40–54 months total construction * **Medium Facility (100–400 beds, 100,000–350,000 sq ft):** 60–84 months total construction * **Large Facility (400+ beds, 350,000+ sq ft):** 100–120 months total construction #### **Other Projects incl. Stadiums, Sports Facilities etc. Construction Timeline** These large-scale venues are unique mega projects combining complex structural, architectural, and mechanical systems with large occupancies. **Timeline by Size** * **Small Facility (<30,000 seats / <100,000 sq ft):** 18–24 months total construction * **Medium Facility (30,000 – 60,000 seats / 100,000 – 300,000 sq ft):** 24–36 months total construction * **Large Facility (60,000+ seats / 300,000+ sq ft):** 36–48+ months total construction # Project Likelihood Thousands of industrial projects are announced each year in the United States, but only a fraction advance beyond early planning. This requires a clear framework for assessing which announced projects are most likely to move forward. To support this, we use an expected project likelihood methodology to evaluate the probability of each project advancing from announcement to construction and completion. Since not every announced project proceeds as planned, each development is assigned a **likelihood rating** based on its progress, visibility, and maturity. Projects are categorized into three levels: **“High” likelihood** – Projects are highly likely to proceed due to secured land control, advanced or approved permitting and zoning, confirmed utility access (power, water, transport), committed financing or incentives, and experienced developers. **“Medium” likelihood** – Projects have a moderate chance of advancing due to pending land acquisition, early-stage permitting, uncertain financing or infrastructure, or limited development history. These projects are typically active but not yet confirmed for construction. **“Low” likelihood** – Projects face significant uncertainty due to unresolved land or permitting issues, lack of clear funding or utility access, stalled timelines, or minimal public or corporate updates since announcement. Using standardized likelihood ratings allows for consistent tracking of project momentum over time. This approach helps identify which developments are gaining traction and which may stall, creating a data-driven view of real industrial progress across states and sectors.
# Data Quality Aterio applies a structured validation process to ensure that all project information such as facility type, construction status, developer details, and location is accurate and consistent with verified public records. Each data point is cross-checked using multiple sources, including corporate filings, press releases, permitting databases, infrastructure filings, and satellite imagery. When no figures are provided, only observable or verifiable information is included in the dataset. Our analysts reconcile project counts and footprints using a combination of regulatory data, developer updates, and visible construction evidence. Each facility is validated individually to ensure that the dataset reflects real, verifiable industrial activity rather than speculative or duplicate announcements.
# Use Cases #### Real Estate Development Provide real estate developers, site selectors, and brokers with early visibility into infrastructure and commercial activity that signals rising demand for land and property. By tracking new data centers, industrial facilities, and large-scale construction projects as they break ground, professionals can identify high-growth regions and emerging clusters before they appear in traditional datasets. This early insight enables smarter decisions around land acquisition, zoning engagement, leasing strategies, and speculative development. Whether targeting infill opportunities near major builds or planning greenfield investments, real estate professionals can act with precision based on real-time activity. #### Market & Economic Analysis Economic development agencies, regional planners, and research firms can use these real-time signals to evaluate industrial growth patterns and their downstream effects on local economies. By analyzing where new factories, data centers, and commercial hubs are emerging, analysts can assess regional competitiveness, forecast job creation, and anticipate infrastructure strain or investment needs. The dataset supports dynamic modeling of workforce demand, supply chain shifts, and tax base expansion—allowing decision-makers to align public resources and private investments with the pace of actual development activity, rather than relying on lagging indicators. # FAQs # What types of projects are included in the dataset and what are excluded? The dataset focuses exclusively on large-scale industrial developments in the United States, covering manufacturing facilities, major energy projects (such as solar farms, power plants), large distribution or logistics hubs, large scale stadiums, arenas, hospitals, and other similarly sized institutional or heavy-industrial builds. Small-commercial developments (e.g., local retail centers, restaurants), regular parks, and residential housing are excluded. The intention is to capture industrial scale announcements that reflect meaningful capital investment, infrastructure build-out. By limiting to these large-scale projects, the dataset avoids noise from small announcements and ensures focus on developments that have the potential to shift infrastructure, land use and utility load patterns. # How up-to-date is the dataset, and how are changes or cancellations handled? The dataset is continuously updated as new announcements, permit filings, imagery and utility/infrastructure data become available. Each record includes timestamps for key milestones (announcement date, construction start date, estimated activation date, etc.) and discretionary fields for cancellation/withdrawal dates where applicable. If a project is formally cancelled or withdrawn, the dataset records the withdrawal date to distinguish it from projects still in play. Because the dataset relies on public and observable sources, each record is refreshed as soon as verified information surfaces. # What do the “likelihood” ratings mean and how should we interpret them? Each project record is assigned one of three likelihood ratings - High, Medium or Low, based on its maturity, visibility and risk profile. A “High” rating indicates that land control is secured, permitting is advanced or approved, utility/infrastructure access is confirmed (power, water, transport), financing or government incentives are committed, and the developer has relevant experience. “Medium” signals that there is progress but more uncertainty remains, such as pending land acquisitions, early-stage permits, or unconfirmed infrastructure access. “Low” means substantial uncertainty exists: perhaps land or zoning is unresolved, infrastructure access is unclear, financing or partner commitments are missing or the developer lacks a proven track record. These ratings allow users to filter or weight the dataset by execution risk rather than taking all announcements as equally likely. # How can I access additional context or background on specific projects? Each record in the dataset includes a Record Notes field, where Aterio’s analysts document any relevant context, findings, or observations that add value to the project’s profile. These notes may include details from local hearings, environmental filings, community discussions, financing updates, or related infrastructure activity. The Record Notes field provides qualitative insight that complements the structured data, helping users understand nuances such as developer intent, project history, or early indicators of progress or delay. # Early Signals on New Developments *Aterio's Early Signals* *on New Developments* product enables AI agents and enterprise systems to respond instantly to physical infrastructure activity across the U.S. With standardized, machine-readable signals on new data centers, factories, and commercial construction updated hourly, this dataset powers automated workflows and business development at scale. By transforming fragmented, location-based development activity into structured, real-time signals, our product bridges the gap between the physical and digital worlds. Whether it's a new data center breaking ground or a commercial facility entering construction, these updates act as live indicators of economic intent—fueling systems that need to act, adapt, or respond as the built environment evolves. This continuous stream of intelligence empowers businesses to anticipate demand, allocate resources efficiently, and stay ahead of market shifts as they happen. # Dataset Explore real-time signals for commercial developments, data centers, and factories, across the U.S. Access the dataset via API, Webhook, JSON Feed, or downloadable formats including CSV and EXCEL—ready to power AI agents, CRMs, and enterprise automation workflows. ### Dataset Components


Real-time event signals for	New construction starts Data center and industrial site announcements Energy infrastructure projects Commercial real estate developments
Signal metadata	Event type and summary Company name and ticker Geographic location (lat/lon, city, state) Timestamp and source reference

### Schema Overview

Field Name	Field Data Type	Description
ATERIO_FEED_UNIQUE_ID	String	Aterio unique ID of the development signal data feed
COMPANY_NAME	String	Associated company name.
COMPANY_TICKER_NAME	String	Company Ticker
SUMMARY	String	Summary with the development details
ZIP_CODE	String	US Zip Code
CITY_NAME	String	City name
STATE_CODE	String	State code
LOCATION_LATITUDE	Float	Latitude coordinate location
LOCATION_LONGITUDE	Float	Longitude coordinate location
REFERENCE_URL	String	URL with the development reference
RECORD_CREATED_DATE	Date	Date the record was created in the system.
RECORD_UPDATED_DATE	Date	Date the record was last modified.

# Data Access Our Real Time Development Signals data product is available on the following platforms:


AWS	S3 Bucket	data-centers-access-aws.png
Snowflake Data Cloud	Data Listing	data-access-snowflakepng.png
Aterio's Download Portal	Customers Only	data-access-intenral.png
Databricks		databricks-logo-access.png
GCP	Cloud Storage & BigQuery	gcp-logo-access.png

[Contact us](https://www.aterio.io/contact-us) to schedule a call and learn more about access options (including APIs/Webhooks) and pricing details. # Data Sources # Business Applications ### AI/LLM teams building real-world automation and copilot tools Real Time Development Signals offer AI and LLM teams high-frequency, structured data on new physical infrastructure projects—essential for grounding AI agents in real-world changes. With hourly updates on data centers, factories, and commercial sites, developers can build copilots that proactively surface opportunities, trigger alerts for relevant activities, or automatically initiate workflows in CRM and ERP systems. This dataset enhances the contextual awareness and responsiveness of automation tools designed to act on real-world developments. ### Territory-based sales intelligence and lead routing Sales teams operating in regional territories can leverage this dataset to identify high-value leads the moment new construction or expansion is signaled. By detecting new data centers, manufacturing sites, or commercial developments as they emerge, sales organizations can route leads to the right reps based on geography and vertical, shorten the sales cycle, and ensure early engagement with projects still in formative stages. The result is smarter lead prioritization and competitive advantage through real-time insight. ### Utility and infrastructure planning For utility providers, engineering firms, and infrastructure planners, these signals offer early awareness of where power, water, broadband, and transportation infrastructure will be needed next. Planners can track the rise of energy-intensive facilities like data centers or factories to model future demand, plan upgrades, and coordinate with local governments. The standardized format and hourly updates help shift planning from reactive to proactive—supporting smarter capital allocation and better regional service delivery. ### Investment and market intelligence platforms This dataset enables investors and analysts to track economic development patterns in near real-time, offering visibility into where capital is flowing and which regions or sectors are accelerating. By monitoring ground activity across commercial, industrial, and technology sectors, platforms can surface early indicators of market momentum, spot emerging clusters, and build predictive models on regional growth. The real-time nature of the signals supports faster, data-driven investment decisions and more dynamic market intelligence products. # US Population Forecast & Economic Development Indicators The US Population Forecast & Economic Development Indicators datasets combine demographic projection with economic measurement at ZIP code level across the United States. It covers historical population from 2010 to the present and forward projections through 2035, broken out by age and gender, with upper and lower bounds published alongside every projection. Alongside the demographic view, the dataset carries the economic indicators that help explain why populations move. These include business establishment counts and their growth, labor force participation, employment and unemployment, and county economic output growth measured overall and across the sectors most closely tied to local development, among them real estate, construction and retail trade. Results are available for individual ZIP codes, with a full monthly history retained so that movement can be traced over time. For researchers, analysts and decision-makers, it offers a single dependable basis for anticipating where demand is forming and understanding the economic conditions driving it. # Main Datasets # Population Forecast by ZIP Code This dataset provides projected population figures at the ZIP Code level, enabling granular analysis of future demographic trends across the United States. It includes total population forecasts based on historical trends, migration patterns, and dynamic factors. Ideal for high-level planning and regional forecasting. ### Schema Overview

Field Name	Field Data Type	Description
ZIP_CODE	String	ZIP Code
CITY_NAME	String	City name
COUNTY_NAME	String	County name
COUNTY_FIPS_CODE	String	Federal Information Processing Standards (FIPS) code for counties
STATE_CODE	String	State Code
TOT_CENSUS_POP_2010	Integer	Population count for Census year 2010
TOT_CENSUS_POP_2011	Integer	Population count for Census year 2011
TOT_CENSUS_POP_2012	Integer	Population count for Census year 2012
TOT_CENSUS_POP_2013	Integer	Population count for Census year 2013
TOT_CENSUS_POP_2014	Integer	Population count for Census year 2014
TOT_CENSUS_POP_2015	Integer	Population count for Census year 2015
TOT_CENSUS_POP_2016	Integer	Population count for Census year 2016
TOT_CENSUS_POP_2017	Integer	Population count for Census year 2017
TOT_CENSUS_POP_2018	Integer	Population count for Census year 2018
TOT_CENSUS_POP_2019	Integer	Population count for Census year 2019
TOT_CENSUS_POP_2020	Integer	Population count for Census year 2020
TOT_CENSUS_POP_2021	Integer	Population count for Census year 2021
TOT_CENSUS_POP_2022	Integer	Population count for Census year 2022
TOT_FX_POP_2023	Integer	Population count for Census year 2023
TOT_FX_POP_2024	Integer	Forecasted population for the year 2024
TOT_FX_POP_2025	Integer	Forecasted population for the year 2025
TOT_FX_POP_2026	Integer	Forecasted population for the year 2026
TOT_FX_POP_2027	Integer	Forecasted population for the year 2027
TOT_FX_POP_2028	Integer	Forecasted population for the year 2028
TOT_FX_POP_2029	Integer	Forecasted population for the year 2029
TOT_FX_POP_2030	Integer	Forecasted population for the year 2030
TOT_FX_POP_2031	Integer	Forecasted population for the year 2031
TOT_FX_POP_2032	Integer	Forecasted population for the year 2032
TOT_FX_POP_2033	Integer	Forecasted population for the year 2033
TOT_FX_POP_2034	Integer	Forecasted population for the year 2034
TOT_FX_POP_2035	Integer	Forecasted population for the year 2035
PCT_CHANGE_PRECOVID	Float	Population variation during Covid
PCT_CHANGE_COVID	Float	Population variation after Covid
PCT_CHANGE_POSTCOVID	Float	Population variation before Covid
UPDATED_AT	Date	Date of the last data update

# Population Forecast by Age & Gender by ZIP Code An enriched version of the population forecast dataset, this file offers projections segmented by age groups and gender at the ZIP Code level. It supports detailed demographic analysis, allowing users to assess age-specific and gender-based population trends for more targeted research, planning, and policy development. ### Schema Overview

Field Name	Field Data Type	Description
YEAR	Integer	Year
ZIP_CODE	String	ZIP Code
COUNTY_NAME	String	County name
COUNTY_FIPS_CODE	String	Federal Information Processing Standards (FIPS) code for counties
CITY_NAME	String	City name
PLACE_FIPS_CODE	String	City FIPS Code
STATE_CODE	String	State Code
GENDER_NAME	String	Gender Name (Male/Female)
AGE_GROUP	String	Age group buckets
TOT_POPULATION	Integer	Total population (Year over year)
TOT_POPULATION_PRED_LOWER	Integer	Total Population Prediction Lower Bound
TOT_POPULATION_PRED_UPPER	Integer	Total Population Prediction Upper Bound
MONTH_CODE	Integer	Month Code (related to UPDATED_AT)
UPDATED_AT	Date	Date of the last data update

# Housing Forecast & Economic Development Indicators A forward view of housing need combined with the local economic conditions that shape it, at the ZIP Code level. It pairs existing housing stock and projected home demand through 2030 with occupancy and vacancy, household income and debt burden, business and employment activity, and prevailing home values and rents. This supports market sizing, site selection and development planning, helping users identify where housing demand is forming and whether local conditions can sustain it. ### Schema Overview | COLUMNS | DATA TYPE | DESCRIPTION | | ------------------------------------------- | --------- | ----------------------------------------------------------------------------- | | ZIP\_CODE | CHAR(5) | US Zip Code | | ATERIO\_UNIQUE\_ID | STRING | Aterio Unique ID for the record | | LOCATION\_LATITUDE | DOUBLE | Latitude coordinate location | | LOCATION\_LONGITUDE | DOUBLE | Longitude coordinate location | | AREA\_LAND\_SQ\_MI | DOUBLE | Land area in square miles | | AREA\_WATER\_SQ\_MI | DOUBLE | Water area in square miles | | PLACE\_FIPS\_CODE | CHAR(5) | City FIPS Code | | CITY\_NAME | STRING | City name | | CBSA\_CODE | STRING | Core-based Statistical Area Code | | CBSA\_TITLE | STRING | Core-based Statistical Area Title | | STATISTICAL\_AREA\_TYPE\_NAME | STRING | Type of CBSA (Metropolitan Statistical Area or Micropolitan Statistical Area) | | COUNTY\_NAME | STRING | County name | | COUNTY\_FIPS\_CODE | CHAR(5) | Federal Information Processing Standards (FIPS) code for counties | | STATE\_CODE | CHAR(2) | State code | | COUNTRY\_NAME | STRING | Country Name | | TOT\_HOME\_AVAILABLE\_2020 | INTEGER | Total available homes in 2020 | | TOT\_HOME\_AVAILABLE\_2021 | INTEGER | Total available homes in 2021 | | TOT\_HOME\_AVAILABLE\_2022 | INTEGER | Total available homes in 2022 | | TOT\_HOME\_AVAILABLE\_2023 | INTEGER | Total available homes in 2023 | | EST\_TOTAL\_HOME\_DEMAND\_2024 | INTEGER | Estimated demand for homes in 2024 | | EST\_TOTAL\_HOME\_DEMAND\_2025 | INTEGER | Estimated demand for homes in 2025 | | EST\_TOTAL\_HOME\_DEMAND\_2026 | INTEGER | Estimated demand for homes in 2026 | | EST\_TOTAL\_HOME\_DEMAND\_2027 | INTEGER | Estimated demand for homes in 2027 | | EST\_TOTAL\_HOME\_DEMAND\_2028 | INTEGER | Estimated demand for homes in 2028 | | EST\_TOTAL\_HOME\_DEMAND\_2029 | INTEGER | Estimated demand for homes in 2029 | | EST\_TOTAL\_HOME\_DEMAND\_2030 | INTEGER | Estimated demand for homes in 2030 | | TOT\_OWNER\_OCCUPIED\_HOUSING\_UNIT | INTEGER | Total owner-occupied housing units | | TOT\_RENTER\_OCCUPIED\_HOUSING\_UNIT | INTEGER | Total renter-occupied housing units | | RT\_VACANCY | DOUBLE | Rate of vacant housing units | | RT\_VACANCY\_RENTAL | DOUBLE | Rate of vacant rental housing units | | RT\_HOMEOWNERSHIP | DOUBLE | Rate of homeownership | | AVG\_HOUSEHOLD\_INCOME\_2021 | INTEGER | US Annual average household income reported in 2021 | | AVG\_HOUSEHOLD\_INCOME\_2022 | INTEGER | US Annual average household income reported in 2022 | | AVG\_HOUSEHOLD\_INCOME\_2023 | INTEGER | US Annual average household income reported in 2023 | | IDX\_AFFORDABILITY | DOUBLE | Index indicating housing affordability | | RT\_DEBT\_INCOME\_2020 | DOUBLE | Debt-to-income rate in 2020 | | RT\_DEBT\_INCOME\_2021 | DOUBLE | Debt-to-income rate in 2021 | | RT\_DEBT\_INCOME\_2022 | DOUBLE | Debt-to-income rate in 2022 | | RT\_DEBT\_INCOME\_2023 | DOUBLE | Debt-to-income rate in 2023 | | RT\_DEBT\_INCOME\_2020\_2021 | DOUBLE | Debt-to-income rate variation 2020-2021 | | RT\_DEBT\_INCOME\_2021\_2022 | DOUBLE | Debt-to-income rate variation 2021-2022 | | RT\_DEBT\_INCOME\_2022\_2023 | DOUBLE | Debt-to-income rate variation 2022-2023 | | PCT\_HOUSEHOLD\_INCOME\_BELOW\_50K\_2020 | DOUBLE | Percentage household debt-to-income below $50K in 2020 | | PCT\_HOUSEHOLD\_INCOME\_50K\_75K\_2020 | DOUBLE | Percentage household debt-to-income $50K-$75K in 2020 | | PCT\_HOUSEHOLD\_INCOME\_75K\_100K\_2020 | DOUBLE | Percentage household debt-to-income $75K-$100K in 2020 | | PCT\_HOUSEHOLD\_INCOME\_100K\_200K\_2020 | DOUBLE | Percentage household debt-to-income $100K-$200K in 2020 | | PCT\_HOUSEHOLD\_INCOME\_ABOVE\_200K\_2020 | DOUBLE | Percentage household debt-to-income above $200K in 2020 | | PCT\_HOUSEHOLD\_INCOME\_BELOW\_50K\_2021 | DOUBLE | Percentage household debt-to-income below $50K in 2021 | | PCT\_HOUSEHOLD\_INCOME\_50K\_75K\_2021 | DOUBLE | Percentage household debt-to-income $50K-$75K in 2021 | | PCT\_HOUSEHOLD\_INCOME\_75K\_100K\_2021 | DOUBLE | Percentage household debt-to-income $75K-$100K in 2021 | | PCT\_HOUSEHOLD\_INCOME\_100K\_200K\_2021 | DOUBLE | Percentage household debt-to-income $100K-$200K in 2021 | | PCT\_HOUSEHOLD\_INCOME\_ABOVE\_200K\_2021 | DOUBLE | Percentage household debt-to-income above $200K in 2021 | | PCT\_HOUSEHOLD\_INCOME\_BELOW\_50K\_2022 | DOUBLE | Percentage household debt-to-income below $50K in 2022 | | PCT\_HOUSEHOLD\_INCOME\_50K\_75K\_2022 | DOUBLE | Percentage household debt-to-income $50K-$75K in 2022 | | PCT\_HOUSEHOLD\_INCOME\_75K\_100K\_2022 | DOUBLE | Percentage household debt-to-income $75K-$100K in 2022 | | PCT\_HOUSEHOLD\_INCOME\_100K\_200K\_2022 | DOUBLE | Percentage household debt-to-income $100K-$200K in 2022 | | PCT\_HOUSEHOLD\_INCOME\_ABOVE\_200K\_2022 | DOUBLE | Percentage household debt-to-income above $200K in 2022 | | PCT\_HOUSEHOLD\_INCOME\_BELOW\_50K\_2023 | DOUBLE | Percentage household debt-to-income below $50K in 2023 | | PCT\_HOUSEHOLD\_INCOME\_50K\_75K\_2023 | DOUBLE | Percentage household debt-to-income $50K-$75K in 2023 | | PCT\_HOUSEHOLD\_INCOME\_75K\_100K\_2023 | DOUBLE | Percentage household debt-to-income $75K-$100K in 2023 | | PCT\_HOUSEHOLD\_INCOME\_100K\_200K\_2023 | DOUBLE | Percentage household debt-to-income $100K-$200K in 2023 | | PCT\_HOUSEHOLD\_INCOME\_ABOVE\_200K\_2023 | DOUBLE | Percentage household debt-to-income above $200K in 2023 | | NUM\_BUSN\_ESTABLISHMENT\_2020 | INTEGER | Total business establishments in 2020 | | NUM\_BUSN\_ESTABLISHMENT\_2021 | INTEGER | Number of business establishments in 2021 | | NUM\_BUSN\_ESTABLISHMENT\_2022 | INTEGER | Number of business establishments in 2022 | | RT\_GROWTH\_BUSN\_ESTABLISHMENT\_2020\_2021 | DOUBLE | Growth rate of business establishments 2020-2021 | | RT\_GROWTH\_BUSN\_ESTABLISHMENT\_2021\_2022 | DOUBLE | Growth rate of business establishments 2021-2022 | | TOT\_POP\_OVER\_16\_LY | INTEGER | Total population over 16 years | | TOT\_POP\_LABOR\_FORCE\_LY | INTEGER | Total population in the labor force | | TOT\_POP\_CIVILIAN\_EMPLOYED\_LY | INTEGER | Total civilian population employed | | TOT\_POP\_CIVILIAN\_UNEMPLOYED\_LY | INTEGER | Total civilian population unemployed | | TOT\_POP\_ARMED\_EMPLOYED\_LY | INTEGER | Total armed population employed | | RT\_ARMED\_EMPLOYED\_LY | DOUBLE | Rate of armed population employed | | RT\_UNEMPLOYMENT | DOUBLE | Rate of unemployment | | AVG\_RT\_UNEMPLOYMENT\_COUNTY\_6M | DOUBLE | Average county unemployment rate over the last 6 months | | RT\_UNEMPLOYMENT\_COUNTY\_LM | DOUBLE | County unemployment rate from the last month available. | | HOME\_VALUE\_INDEX\_ZILLOW\_LM | DOUBLE | Home value index on Zillow last month. | | HOME\_VALUE\_INDEX\_ZILLOW\_6M | DOUBLE | Home value index on Zillow 6 months ago | | HOME\_VALUE\_INDEX\_ZILLOW\_12M | DOUBLE | Home value index on Zillow 12 months ago | | OBSERVED\_RENT\_VALUE\_ZILLOW\_LM | DOUBLE | Observed rent value 6 months ago | | OBSERVED\_RENT\_VALUE\_ZILLOW\_6M | DOUBLE | Observed rent value 12 months ago | | OBSERVED\_RENT\_VALUE\_ZILLOW\_12M | DOUBLE | Observed rent value 12 months ago | | ATERIO\_SCORE | DOUBLE | Aterio scoring metric | | NUM\_VERSION\_ATERIO\_SCORE | STRING | Aterio scoring metric version | | MONTH\_CODE | INTEGER | Month Code | | WEEK\_CODE | INTEGER | Week Code | | UPDATED\_AT | DATE | Date of the last data update | # Population Forecast Methodology # Baseline Model The baseline population forecast model is updated annually, reflecting the latest data releases from our primary sources, such as the US Census. ### **Main Data Sources** * Population Estimates from US Census (County and ZCTA Level) * Fertility Rates from Centers from Disease Control and Prevent (CDC) WONDER database. * Mortality Rates from Centers from Disease Control and Prevent (CDC) WONDER database. * U.S. County Population Data Age/Gender from National Cancer Institute (NCI) ### Baseline Forecast Methodology: Age-Gender Cohort Component Model For this baseline approach, the cohort component model is applied to age and gender groups using the following main formula: $$ P\_{s,a+1,t+1} = P\_{s,a,t} \cdot (1 - M\_{s,a,t}) + N\_{s,a+1,t+1} $$ **Where:** * P_s,a,t: Population of sex s, age a, in year t * M_s,a,t: Mortality rate for sex s, age a, in year t * N_s,a,t: Net migration for sex s, age a, in year t Of course, new births (age 0) require special treatment, calculated based on the fertility rates of women aged 15 to 45 years: $$ P\_{s,0,t+1} = \left( \sum\_{a=15}^{49} P\_{f,a,t} \cdot F\_{a,t} \cdot S\_s \right) \cdot (1 - M\_{s,0,t}) + N\_{s,0,t+1} $$ **Where**: * F_a,t: Fertility rate for females aged a in year t * S_s: Proportion of births of sex s (e.g., 0.512 for male, 0.488 for female) For migration patterns, we incorporate annual indicators that help guide the direction of population movement (inflows and outflows), such as: * Housing market trends * Employability * And others ### State-Level Calibration After building our cohort component model, we saw a chance to improve our projections by comparing them with state-level data from respected research groups. This comparison gave us important insights into regional population differences and helped validate the strength of our initial model. Using these insights, we added a calibration step that incorporated annual projection data from key states like California and Texas. By adjusting our model to match these expert projections, we enhanced the accuracy and reliability of our forecasts, providing results that are both data-driven and backed by professional expertise. # Dynamic Model Our population forecast model operates at the ZIP Code level and begins with a robust cohort-component baseline. This baseline is then calibrated using the most recent regional trends to better reflect real-world dynamics and provide more precise, localized projections. **Data Sources Integrated into the Model:** * **Housing Market Trends**: Including new listings, housing permits, and housing turnover. * **Housing Values**: Changes in median home prices and rental costs as proxies for economic pressure and demand. * **Employability Metrics**: Local job availability, unemployment rates, and job growth indicators. * **Estimated Audiences (Meta)**: Aggregated behavioral and location-based data to assess population movement and digital footprint. ### Generation Dynamic Signals by ZIP Code This component tracks shifts in the distribution of generational cohorts across ZIP Codes. Using age breakdowns from census-based sources and historical cohort patterns, we monitor how these groups evolve spatially over time. This signal captures short-term shifts in the age composition of local populations, particularly in response to economic conditions. By analyzing monthly or annual changes in population estimates by age group, we can detect emerging patterns—such as younger working-age cohorts moving into high-opportunity areas or leaving ZIP Codes experiencing economic pressure. Rather than focusing on long-term life-cycle trends like retirement or family formation, our approach emphasizes more immediate drivers such as **housing affordability**, **job availability**, and **income trends**. For example, a sudden rise in unemployment or decline in housing inventory may correspond with out-migration among younger renters or working-age individuals. These signals help us adjust projections at the ZIP Code level by flagging areas where **short-term in- or out-migration** is likely occurring among specific age groups. When combined with other indicators such as employment data and housing trends, this allows us to produce more responsive, localized forecasts that reflect real-time economic dynamics. ### Development-Driven Population Boost In addition to demographic and economic signals, our model incorporates a secondary adjustment layer that accounts for **planned or ongoing developments** within each region. These include large-scale projects such as **data centers, energy infrastructure, logistics facilities and others**. When such developments are identified near a ZIP Code or county, we apply a population “boost” based on the expected impact these projects may have on **job creation**, **in-migration**, and **housing demand**. These forward-looking signals help us anticipate population changes that are not yet visible in historical data but are likely to materialize in the near future due to increased economic activity. This enhancement allows the model to be more **proactive** rather than purely reactive—capturing areas of future growth earlier, and improving the precision of our projections where economic momentum is already forming on the ground. # Aterio Indices & Score # Methodology ### Overview Aterio scores every United States ZIP code on its underlying residential real estate fundamentals. It is designed to answer one practical question at national scale: where are the conditions for residential investment strongest, and why? The assessment rests on six independent dimensions: population momentum, the balance between housing demand and supply, household financial capacity, economic development, downside risk, and investment opportunity. Each is scored separately, and the six are then combined into a single headline measure, the Aterio Score. Both the headline score and the six underlying dimensions are delivered together with the evidence behind them, so that every result can be understood and challenged rather than taken on trust. A ZIP code does not simply score well. It scores well for reasons that are visible in the data alongside it. ### How the assessment works #### A comparative measure Every dimension positions a ZIP code against the full national distribution. A high score means the area performs strongly relative to the rest of the country on that dimension, rather than having crossed a fixed threshold. This makes the measures well suited to screening, ranking and shortlisting across large geographies. #### Consistent direction Five of the six dimensions are oriented so that a higher value is more favourable. The exception is Risk Factors, where a higher value indicates greater exposure. When the six are combined into the headline Aterio Score, the risk dimension is inverted, so the headline reads consistently: higher is stronger. #### Built for comparison, not prediction The assessment describes the fundamentals of a market as they stand, and the trajectory those fundamentals imply. It is a framework for comparing places rather than a forecast of returns on any individual asset. ### 3. The six dimensions #### 3.1 Population and Migration What it answers: *Is this market growing, and how much of it is there?* What it measures. Demographic momentum: the scale of the resident population today and, more importantly, the direction and pace of its projected change over the coming decade. Both matter, but trajectory is weighted more heavily than size. A smaller market on a strong upward path will score above a larger one that has stalled. What informs it. Decennial census counts, Aterio's proprietary population forecast running through 2035, county-level net migration statistics, and independent academic population projections used as a cross-check. How to read it. High values identify areas gaining residents. Because population growth precedes housing demand, this dimension is the leading indicator in the framework, and it carries the greatest influence on the headline score. #### 3.2 Demand and Supply What it answers: *Will there be enough housing here?* What it measures. The balance between the housing that will be needed and the housing that exists. Projected population is translated into projected housing need using the way households in that specific area actually occupy homes, then compared against the current stock. What informs it. Census housing inventory, local household occupancy patterns, residential building permit activity, and Aterio's population forecast. Vacancy and owner-occupancy measures are reported alongside as supporting context. How to read it. High values indicate a projected shortfall, with demand outpacing available stock. These are the markets where supply pressure tends to translate into pricing power for owners and into opportunity for developers. Low values indicate an area already well supplied relative to what its population is expected to need. #### 3.3 Capacity to Pay What it answers: *Can households here afford to live here?* What it measures. The financial strength of local households relative to local housing costs. It brings together how much households earn, how that income is distributed across the area, how much debt households carry, and how local home values stand in relation to local incomes. What informs it. Census household income levels and income distribution, Federal Reserve household debt-to-income statistics, and prevailing local home values. How to read it. High values indicate residents who can comfortably support housing costs, which supports both price stability and reliable rent collection. Low values flag markets where housing costs have moved ahead of local earning power, a stress signal even where prices are currently rising. #### 3.4 Economic Development What it answers: *Is the local economy outperforming the country?* What it measures. The strength of local economic output growth measured against the national average, with particular emphasis on real-estate-linked activity rather than the economy as a whole. The dimension is deliberately relative: it rewards areas pulling ahead of the national trend. What informs it. Bureau of Economic Analysis county economic output statistics, overall and by sector. Business establishment counts, labour force participation and employment measures are reported alongside as supporting context. How to read it. High values indicate a local economy expanding faster than the country as a whole, particularly in the sectors most closely tied to property. Because the underlying statistics are published at county level, this dimension describes the wider economic region a ZIP code sits within. #### 3.5 Risk Factors What it answers: *What could go wrong here?* What it measures. Downside exposure across the conditions that most directly affect whether residents stay, whether tenants pay, and whether property retains its value: local labour market weakness measured against the long-run national norm, violent and property crime, and the quality and safety of the public drinking water supply. What informs it. Bureau of Labor Statistics county unemployment data, FBI crime statistics, and Environmental Protection Agency drinking water compliance records. How to read it. This dimension runs in the opposite direction to the others. A higher value means greater risk, and it is inverted before it contributes to the headline score. Alongside it, the dataset separately reports natural hazard exposure, expected annual loss, community resilience, social vulnerability, flood insurance coverage and cost, mortgage default history and home ownership rates, each available as its own measure so that clients can apply their own risk lens. #### 3.6 Opportunity What it answers: *Does the income justify the price?* What it measures. The investment return characteristics of the market: how prevailing local rents compare with the cost of financing a typical home at current mortgage rates, adjusted for how stable local home values have been over the preceding two years. Yield carries the greater weight, with price stability moderating it. What informs it. Zillow home value and observed rent indices, prevailing thirty-year mortgage rates, and two years of local price history. Federal fair market rents and census rent measures are reported alongside as independent reference points. How to read it. High values indicate markets where rental income covers financing comfortably and where values have held steady. Low values indicate either thin yield or an unstable price base. This dimension depends on sufficient local market activity to be measured reliably. Where that activity is too sparse, typically in very rural areas, it is not reported, and the headline score is built from the remaining five dimensions instead. ### 4. The Aterio Score The Aterio Score brings the six dimensions together into a single headline measure. It is a weighted combination rather than a simple average, because the dimensions do not carry equal predictive value. Population momentum carries the greatest influence, reflecting the view that demographic trajectory is the most durable driver of long-term residential demand. Housing supply balance and downside risk follow closely. Household financial capacity, economic development and investment yield each carry a lighter but material influence.

Dimension	Influence on the headline score	Direction
Population and Migration	Greatest	Higher is stronger
Demand and Supply	Substantial	Higher is stronger
Risk Factors	Substantial	Lower is stronger
Capacity to Pay	Moderate	Higher is stronger
Economic Development	Moderate	Higher is stronger
Opportunity	Supporting	Higher is stronger

Where a market has insufficient pricing activity for the Opportunity dimension to be measured, its influence is redistributed across the remaining dimensions rather than treated as a zero. This keeps the headline score available and meaningful in rural and low-transaction markets while remaining transparent, because the dimension itself is shown as unavailable rather than filled in. Population momentum is the one dimension treated as indispensable. Where an area's demographic trajectory cannot be established with confidence, typically in very small or unusual ZIP codes, no headline score is published for it. Aterio would rather report nothing than report a figure it does not stand behind. ### 5. The data behind the assessment Aterio builds on established, citable sources rather than opaque inputs. The great majority of the assessment rests on official United States government statistics, supplemented by leading commercial market data and Aterio's own forecasting work. #### Official statistics

Source	Contribution
US Census Bureau	Population counts, household income and income distribution, housing inventory and occupancy, home ownership and vacancy, labour force and employment, rent levels, building permits, business establishments
Bureau of Labor Statistics	County unemployment
Bureau of Economic Analysis	County economic output growth, overall and by sector
Federal Reserve	Household debt-to-income, mortgage rates, net migration
Federal Bureau of Investigation	Violent and property crime
Environmental Protection Agency	Public drinking water compliance and contamination
FEMA	Natural hazard risk, expected annual loss, community resilience, social vulnerability, flood insurance coverage and cost
Housing and Urban Development	Fair market rents
Federal Housing Administration	Mortgage default history

#### Market data

Source	Contribution
Zillow	Home value and observed rent indices, with multi-year history
Realtor.com	Listing prices, days on market, listing and pending inventory, price movement and market activity measures

#### Aterio proprietary

Source	Contribution
Aterio population forecast	ZIP code level population projections through 2035
Aterio climate resiliency	Composite climate resiliency assessment

Independent academic population projections are also used as a cross-check on Aterio's own forecast where local coverage requires it. ### 6. What is delivered Alongside the headline score and the six dimensions, each ZIP code record carries the supporting evidence, so that any result can be examined directly. * Population and demographics. Resident population from 2010 to the present and Aterio's projection through 2035, with growth rates over standard intervals and across the pre-pandemic, pandemic and post-pandemic periods. * Housing. Housing inventory, projected housing need, the resulting balance, owner and renter occupancy, vacancy rates and building permit activity. * Household finances. Average household income, the distribution of households across income bands, household debt burden, and the relationship of local home values to local incomes. * Economy and employment. Business establishment counts and their growth, labour force size, employment and unemployment, and county economic output growth by sector. * Pricing and yield. Home values and observed rents at current, six-month and twelve-month points, price direction and rate of change, momentum and volatility measures, an indicative financing cost, and the resulting relationship between rent and financing. * Rents. Federal fair market rents and census rent levels, each broken out by property size. * Risk. Unemployment, crime rates and their standing against the national average, drinking water quality, natural hazard exposure, expected annual loss, community resilience, social vulnerability, flood insurance coverage and cost, mortgage default history and home ownership. * Live market activity. Listing prices and price per square foot, days on market, new, active, pending and reduced listings, and market activity measures, each with month-on-month and year-on-year movement and comparison against the national picture. * Geography. Coordinates, city, county, metropolitan area and state, land and water area, and population density. ### 7. Coverage and refresh

Aspect	Detail
Geography	United States, one record per ZIP code
Refresh	Monthly
Market data	Home values, rents, listing activity and unemployment refresh monthly
Official statistics	Refresh on each publishing authority's own schedule, typically annually
Versioning	Every record carries the methodology version and the date it was produced, so results can always be traced to the basis on which they were calculated
History	A monthly historical record is retained

Because official statistics are published on annual cycles, the dimensions built on them are stable by design. Month-to-month movement in the headline score comes principally from live market pricing and labour market conditions. This is the intended behaviour, since structural fundamentals should not oscillate monthly. ### 8. Important notes * The measures are comparative. The dimensions position each ZIP code against the national distribution as it stands at the time of production. They are designed for comparing markets with one another rather than for measuring against a fixed absolute standard. * Compare within an edition. Because each edition is calculated against the national picture at that point in time, scores are directly comparable within an edition. Comparison across methodology versions should account for the version recorded against each result. * Coverage varies by market. Some measures require a minimum level of local activity or population to be reported reliably. Where that threshold is not met, the measure is shown as unavailable rather than estimated, and that availability is visible in the data itself. * The financing figure is indicative. The financing cost reported alongside the Opportunity dimension is an indicative reference based on prevailing national mortgage rates and the local home value. It is intentionally simplified and does not represent a quotation, an underwriting decision, or the full cost of ownership. * This is not investment advice. Aterio Investment Insights is a market analysis product. It describes market fundamentals and is not investment, financial, legal or tax advice, and it is not a recommendation regarding any specific property or transaction. Users should apply their own diligence and judgement. ### 9. Glossary

Term	Meaning
Aterio Score	The headline measure, combining all six dimensions into a single comparative assessment of a ZIP code's residential investment fundamentals. Higher is stronger.
Dimension	One of the six independent components of the assessment, each scored and reported in its own right.
ZIP code	The geographic unit of the assessment. Each ZIP code is assigned to a single county, city and metropolitan area so that regional measures attach consistently.
Momentum	A measure of how sustained and one-directional recent price or rent movement has been, distinguishing steady trends from short-lived swings.
Volatility	A measure of how much local values have fluctuated over the preceding two years. Lower volatility indicates a more stable market.
Fair market rent	A federally published rent reference by property size, used as an independent benchmark alongside observed market rents.
Metropolitan area	The wider statistical region a ZIP code belongs to, used to place local results in regional context.

# Dataset

COLUMN	DATA TYPE	DESCRIPTION
ZIP_CODE	CHAR(5)	US Zip Code. Primary key of the dataset
CITY_NAME	STRING	City name associated to the ZIP code
COUNTY_NAME	STRING	County name associated to the ZIP code
COUNTY_FIPS_CODE	CHAR(5)	County FIPS code associated to the ZIP code
STATE_CODE	CHAR(2)	State code associated to the ZIP code
IDX_POPULATION_MIGRATION	DOUBLE	Population and Migration index
IDX_DEMAND_SUPPLY	DOUBLE	Demand and Supply index
IDX_ECONOMIC_DEV	DOUBLE	Economic Development index
IDX_CAPACITY_PAY	DOUBLE	Capacity to Pay index
IDX_RISK_FACTOR	DOUBLE	Risk Factors index. Higher values indicate greater risk
IDX_OPPORTUNITY	DOUBLE	Opportunity index
ATERIO_SCORE	DOUBLE	Aterio Score. Weighted composite of the six dimension indices
MONTH_CODE	INT	Month of the data release. Format YYYYMM
UPDATED_AT	TIMESTAMP	Date the record was produced

# Data Access Our Population Forecast data product is available on the following platforms:


AWS	S3 Bucket	data-centers-access-aws.png
Snowflake Data Cloud	Data Listing	data-access-snowflakepng.png
Aterio's Download Portal	Customers Only	data-access-intenral.png
Databricks		databricks-logo-access.png
GCP	Cloud Storage & BigQuery	gcp-logo-access.png

[Contact us](https://www.aterio.io/contact-us) to schedule a call and learn more about access options and pricing details. # Use Cases ### Market Analysis Population forecasts enable companies to identify emerging markets by analyzing future demographic trends. By understanding shifts in population size, age distribution, and regional growth, businesses can tailor their marketing strategies and product offerings to meet the anticipated demands of different population segments. For instance, a projected increase in the elderly population might prompt healthcare companies to expand their range of senior care products and services. ### Investment Planning Investors use population and migration trends to make strategic investment decisions. Demographic forecasts provide insights into potential growth areas, helping investors allocate resources effectively. For example, a significant influx of people into urban areas can signal opportunities in the real estate and retail sectors. Understanding these trends allows investors to anticipate market needs and reduce investment risks. ### Housing Market Planning Developers rely on population forecasts to anticipate future housing demand. By analyzing demographic projections, developers can plan residential and commercial real estate projects that align with population growth patterns. This proactive approach helps in preventing housing shortages and ensures that infrastructure developments meet the needs of future residents. ### Urban Planning Local and national governments use population forecasts to plan for infrastructure development, housing, transportation, and other urban amenities. Accurate population projections enable urban planners to design cities that accommodate future growth, ensuring sustainable and well-organized urban environments. This includes planning for schools, hospitals, public transport systems, and recreational facilities. ### Resource Allocation Hospitals and healthcare providers utilize demographic forecasts to plan for future healthcare needs. Understanding population growth and aging trends helps in allocating resources, such as medical staff, facilities, and equipment, to areas where they will be most needed. This foresight ensures that healthcare systems are prepared to meet the demands of a changing population. ### Risk Assessment Population forecasts assist insurance companies in assessing risks related to property, health, and life insurance. By analyzing demographic trends, insurers can predict potential changes in risk profiles and adjust their policies accordingly. For example, an aging population may lead to higher demand for health and life insurance products, while urbanization trends might impact property insurance risks. # Introduction Aterio data is built to drop into the tools and workflows your team already uses. Whether you work in a cloud data warehouse, pull data through code, or want it available to your AI tools, we deliver it in a format that fits. ### What you're working with Aterio isn't a set of flat lists. It's a connected view of the infrastructure landscape, where projects, the companies behind them, and the links between them are all mapped. • Every project shows who's behind it: developers, asset owners, investors, and parent companies, labeled by role. • Every company connects the picture, showing what it's involved in across data centers, industrial, and power generation. • Products link to each other: power generation connects to the data centers it powers, including behind-the-meter supply. • The data has history: projects carry lifecycle events and are available point-in-time, so you can see how the landscape changes. ### How access works Access to Aterio data is provisioned as part of onboarding — it is not self-serve. We scope each integration to the products you're subscribed to and set it up with you directly. This keeps your access secure and matched to your use case. If you'd like to explore the data or see how a delivery option fits your stack, the best next step is a short call. ### Which delivery option is right for you A quick guide to the option that usually fits best. If you're not sure, we'll help you pick on the call.

If you want to…	Best fit
Query Aterio data next to your own, with no pipelines to run	Databases
Build a product or automate a sync on your own schedule	API
Make the data reachable by your AI tools and agents	MCP
Grab ready-to-use files manually, no technical setup	Download portal

*Each product can be delivered through any of the options in this section* # Data Warehouses & Lakes Aterio data delivered directly into your cloud data warehouse and data lake, sitting natively alongside your own data with no pipelines to build or maintain. We support the major platforms, so the data lands wherever your team already works: Amazon S3, Databricks, Snowflake, and BigQuery.

{% hint style="info" %} **Using a different data lake or warehouse?** We support S3, Databricks, Snowflake, and BigQuery out of the box, but we're not limited to them. If your team works somewhere else, let us know! {% endhint %} ### When this is a good fit • Your analytics and data teams live in a warehouse day to day • You want Aterio data queryable next to your internal datasets • You'd rather not build or run ingestion pipelines • You need governed, centralized access for a whole team ### What you can do with it • Query the full dataset with the tools you already use • Join Aterio data against your own data for custom analysis • Trace connections across products, like the power behind a data center or every project tied to a company • Feed it into dashboards, models, and internal reporting ### How access works We share the data into your warehouse and scope it to the products you're subscribed to. Setup is handled with you during onboarding. *** **Ready to see the data?** A short call is the best way to see it in action and find the setup that fits. [Book a call →](https://www.aterio.io/contact-us) # Download Portal Access ready-to-use files through a secure portal, with no technical integration required. The simplest way to get Aterio data in hand. Built for teams that want the data without building or running anything.

### When this is a good fit • You want the data without a technical integration • You prefer to grab files manually when you need them • You're evaluating the data or working with it in your own tools ### What you can do with it • Download the products you're subscribed to as ready-to-use files • Work with the data in whatever tools your team already uses • Get started quickly, with no setup on your side ### How access works Portal access is provisioned during onboarding and scoped to your subscribed products. *** **Ready to see the data?** A short call is the best way to see it in action and find the setup that fits. [Book a call →](https://www.aterio.io/contact-us) # Application Programming Interface (API) Pull Aterio data programmatically into your own applications, systems, and workflows, on your own schedule with the full connected dataset behind it, not just flat records. There are two ways to use it, and you can mix both: • Query it live for specific projects, companies, and the connections between them — ideal for lookups and targeted slices. • Take the whole dataset as a single export when you just want the complete product to load into your own environment, with no filtering to set up.

### When this is a good fit • You're building a product or internal tool on Aterio data • You want to sync data into your own systems automatically • You want the entire dataset in hand without querying or filtering • Your team prefers code-level control over how data flows ### What's available through it Across all four products, the API works over the same connected data: • Projects (Data Centers, Power Generation, Industrial Developments): each with its stage, location, and the companies attached to it by role: developer, owner, investor, parent • Companies: each with its industry classification and every project it's tied to, across all products • Relationships: the links between projects, companies, and products come back as part of the data, not something you reconstruct yourself You can retrieve the latest data or an earlier point-in-time version to see how the landscape looked at a past date. ### Take the whole dataset If you don't need querying or filtering, you can pull an entire product as a complete dataset export and work with it in your own environment. Take the latest version, or an earlier point-in-time version. Best for teams that want the full picture in one place and their own tools to slice it. ### How access works API access is provisioned during onboarding and scoped to your subscribed products. *** **Ready to see the data?** A short call is the best way to see it in action and find the setup that fits. [Book a call →](https://www.aterio.io/contact-us) # Model Context Protocol (MCP) Make Aterio data available directly to your AI tools and agents through the Model Context Protocol, so they can query it as part of a workflow, with the full connected dataset behind it. Built for teams working in AI-native environments who want infrastructure and market data reachable by their assistants. • Your team works with AI assistants and agents day to day • You want Aterio data usable inside those workflows, not copied around • You're building agentic or AI-driven analysis on top of our data

### What you can do with it • Let your AI tools pull Aterio data on demand during a task • Ask questions that follow the connections: the companies behind a project, or the power supplying a data center • Ground assistant answers in current, structured data instead of guesses • Build agent workflows that reason over our datasets across products ### How access works MCP access is set up with you during onboarding and scoped to your subscribed products. *** **Ready to see the data?** A short call is the best way to see it in action and find the setup that fits. [Book a call →](https://www.aterio.io/contact-us) # Product and Methodology FAQs ### What data products does Aterio offer? Aterio provides datasets covering US Data Centers, US Industrial Developments, US Power Generation, Balancing Authority Energy Demand, and Population and Housing Forecasts. These products are designed to help investors, utilities, developers, and industrial companies understand where future energy demand and supply are likely to emerge. ### What types of data center projects are included? The US Data Centers Dataset includes operational, under-construction, announced, delayed, cancelled, withdrawn, and land-bank developments. Projects are tracked at the building level, with campus-level information included where multiple buildings are part of the same development. ### What information is available for each data center? Available fields may include provider, developer, project stage, construction progress, estimated activation date, construction dates, location, acreage, building area, published and estimated power capacity, utility, balancing authority, substation, generators, tenants, source links, and satellite imagery references. Field availability depends on the amount of reliable public information available for each project. ### How does Aterio estimate data center power capacity? Aterio prioritizes publicly disclosed power capacity from developers, utilities, regulators, permits, and interconnection documents. When no reliable figure is disclosed, Aterio may estimate total facility power using building size, building type, equipment layouts, satellite imagery, comparable facilities, utility infrastructure, and proprietary power-density models. Aterio’s estimated power generally represents total facility demand, including IT equipment, cooling, and supporting infrastructure, rather than critical IT load alone. ### What is the difference between published power and Aterio-estimated power? Published power is a capacity figure disclosed by a developer, utility, regulator, permit, or another public source. Aterio-estimated power is calculated when no sufficiently reliable public figure is available. The estimate is based on available project characteristics, satellite analysis, comparable developments, and Aterio’s proprietary methodology. Published and estimated values are maintained separately so users can distinguish between sourced and modelled capacity. ### How does Aterio determine construction start dates? Construction start dates are based on evidence of meaningful physical activity, such as site clearing, grading, foundation work, structural construction, or public confirmation of groundbreaking. Preliminary surveying, fencing, soil testing, or minor site preparation may not be treated as full construction. When the exact date is unavailable, Aterio assigns an estimated date based on the earliest confirmed evidence. ### How does Aterio determine data center activation dates? A data center is considered activated when there is evidence that the building has entered service or is ready to support operational workloads. Signals may include installed electrical and cooling equipment, completed parking and access areas, removal of heavy construction equipment, commissioning activity, utility energization, tenant confirmation, or public statements. Construction completion may occur before activation because testing and commissioning can continue after major physical work is finished. ### When are estimated dates revised? Estimated dates are revised when new evidence indicates that the original timeline is no longer realistic. This may include limited construction progress, permitting delays, utility interconnection changes, financing issues, redesigned project scope, developer announcements, satellite-observed inactivity, or changes to equipment-delivery schedules. Aterio does not automatically retain an outdated announced date when project evidence points to a delay. ### What is included in the Industrial Developments dataset? The Industrial Developments dataset tracks major US projects across sectors such as manufacturing, semiconductors, logistics, chemicals, energy-intensive industry, commercial development, housing, healthcare, and other large developments. Fields may include project name, company, location, development stage, capital investment, acreage, building size, permits, project dates, public sources, and execution likelihood. ### What is included in the Power Generation dataset? The US Power Generation Dataset covers operational, under-development, and planned generation infrastructure. Project types may include solar, wind, battery storage, natural gas turbines, reciprocating engines, hydrogen, combined heat and power, distributed generation, transmission infrastructure, and other generation technologies. Available fields may include technology, capacity, developer, location, project stage, construction progress, interconnection status, balancing authority, utility, onsite or grid-connected status, and expected operating dates. ### What does Behind-the-Meter mean? Behind-the-Meter, or BTM, refers to power generation located at or directly connected to the customer’s site and intended to serve onsite electricity demand. BTM generation may reduce the amount of electricity drawn from the wider transmission or distribution system. Examples can include onsite natural gas generation, solar, batteries, fuel cells, or combined heat and power systems. A project with a power purchase agreement is not necessarily BTM. A conventional PPA may involve electricity generated elsewhere and delivered through the grid. ### How does Aterio identify Behind-the-Meter generation? Aterio assigns the BTM classification when public information indicates that generation is located onsite, directly connected to the load, or specifically developed to serve a facility without relying entirely on conventional grid delivery. Sources may include permits, developer disclosures, utility filings, equipment specifications, site plans, environmental applications, and satellite imagery. The field is only populated when sufficient evidence is available. ### How does Aterio assess project execution likelihood? Execution likelihood reflects the strength of evidence that a project will proceed. Relevant signals may include land ownership or site control, permitting progress, utility agreements, interconnection activity, financing, tenant confirmation, construction activity, equipment procurement, developer experience, and public commitments. Projects with confirmed construction, financing, or major customers will generally receive a stronger likelihood assessment than projects supported only by an early announcement. ### What sources does Aterio use? Aterio uses publicly available and commercially licensed sources, including: * Developer and provider websites * Company announcements and press releases * Investor presentations * Regulatory and utility filings * Building permits and zoning documents * Interconnection queues * ISO and RTO information * Environmental permits * Public meeting records * Satellite imagery * Energy-market data * Industry publications ### How is the data validated? Project information is reviewed and structured by Aterio analysts. Material updates are supported by public sources, licensed satellite imagery, or both. Data quality reviews may include source comparison, location validation, duplicate checks, satellite analysis, methodology rules, engineering review, and consistency checks across related buildings, campuses, companies, utilities, and power projects. # Data Delivery and Integration FAQs ### What data delivery options are available? Aterio can deliver data through: * AWS S3 * Google Cloud Storage * Snowflake secure shares * SFTP * APIs * Direct CSV or Parquet files The appropriate delivery method depends on the customer’s infrastructure, preferred file format, update frequency, and internal data-ingestion process. ### Does Aterio provide a customer download portal? Yes. Aterio provides a secure customer download portal where authorized users can access and download subscribed datasets. ### What file formats are supported? Aterio commonly supports CSV, Parquet, and JSON. CSV is suitable for manual analysis and broad compatibility. Parquet is recommended for larger datasets and database workflows because it is compressed and preserves data types. JSON is generally used for API responses and application integrations. Available formats may vary by product and delivery method. ### How frequently is the data updated? Update frequency depends on the product: * US Data Centers: generally updated daily * US Power Generation: generally updated daily or as material changes are identified * Industrial Developments and Early Signals: updates may be delivered hourly * Balancing Authority Energy Demand: updated daily * Population and Housing Forecasts: generally updated monthly Delivery schedules can also be configured based on customer requirements. ### How are AWS S3 deliveries structured? Cloud-storage deliveries can include full dataset snapshots, incremental updates, or both. Files are typically organized using consistent folders, table names, dates, and file-naming conventions. A delivery may contain separate folders or files for inventory, events, reference tables, and historical snapshots. The final structure is agreed during implementation to align with the customer’s ingestion workflow. ### How does Snowflake delivery work? Aterio can provide data through a Snowflake secure share. Customers can access the shared tables directly from their own Snowflake environment without requiring Aterio to send separate files. Customers may query the shared data directly or copy it into their internal databases, depending on their own Snowflake architecture and permissions. ### Does Aterio support API access? Yes. Aterio can provide API access for supported products and use cases. API responses are generally provided in JSON and may support filters, pagination, date parameters, project identifiers, or other query options. Authentication is handled through secure API credentials or tokens. ### Are full snapshots updates available? Yes. Depending on the delivery arrangement, Aterio can provide complete snapshots. A full snapshot contains the latest available version of every included record. Customers that need a simple replacement workflow may prefer full snapshots, while customers maintaining a data warehouse may prefer incremental updates. ### What is the difference between the Inventory and Events tables? The Inventory table contains the latest available record for each project, campus, or building. The Events table records material changes over time. Events may include: * Project announcements * Construction starts * Activation dates * Cancellations or withdrawals * Construction milestones The Events table allows customers to analyze project history and pipeline movement rather than only viewing the current project status. ### How far back does historical data go? Historical coverage varies by product and field. For the Data Centers product, Aterio maintains historical activation information for older operational facilities and detailed daily change history from the point when systematic historical tracking began. Customers should review the relevant product data dictionary because historical availability may differ between construction stages, activation dates, capacity values, and event types. ### How are schema changes communicated? Aterio communicates material schema changes before implementation whenever reasonably possible. Notifications may include the field name, data type, description, effective date, and whether the change is additive or may affect an existing integration. New fields are generally added in a way that minimizes disruption. Customers should build ingestion processes that tolerate additional columns where possible. ### Is a data dictionary available? Yes. Aterio provides data dictionaries for its datasets. The dictionary describes field names, definitions, formats, units, expected values, and relevant methodology. Product-specific documentation may also explain project stages, estimated fields, event types, power calculations, and date logic. ### How is the delivered data secured? Aterio uses the security controls available through each delivery channel. These may include cloud identity and access management, encrypted storage and transmission, Snowflake secure shares, API tokens, restricted SFTP access, one-time-password protection, role-based access, and customer-specific permissions. Access is limited to authorized users and can be removed or modified when requested. ### Can delivery be customized for a customer’s infrastructure? Yes. Aterio can work with customers to configure file format, delivery channel, folder structure, naming conventions, table selection, delivery frequency, and full-versus-incremental update logic. The objective is to align delivery with the customer’s existing database, cloud, analytics, or ingestion environment while maintaining a consistent underlying data model. ### How can the Data Center Inventory and Behind-the-Meter datasets be joined? The Data Center Inventory and Behind-the-Meter datasets can be joined using `ATERIO_POWER_GENERATION_PROJECT_UID`. Because the Behind-the-Meter dataset is structured at the phase level—project, plant, and phase—a direct join may create multiple rows for the same data center record when a project includes several phases. To avoid duplication, first aggregate the Behind-the-Meter data to one row per `ATERIO_POWER_GENERATION_PROJECT_UID`, either by pivoting phases into separate columns or consolidating them into project-level fields. When aggregating: * Sum `TOT_PHASE_NAMEPLATE_POWER_MW` across all phases to calculate the project’s total nameplate capacity. * Do not sum `TOT_CONTRACTED_POWER_CAPACITY_MW`, as this field is already reported at the project level. # Glossary ### Data Centers **Activation Date:** The date a data center building is estimated or confirmed to begin operating and consuming power. **Announced:** A project that has been publicly disclosed but has not yet entered physical construction. **Campus:** A development site containing one or more data center buildings. **Critical IT Load:** The electrical capacity available specifically for servers and IT equipment, excluding cooling and other facility systems. **Data Center Building:** An individual physical facility within a data center campus. **Facility Power / Total Power:** Total electrical demand for the facility, including IT equipment, cooling, lighting, and supporting infrastructure. **Hyperscale Data Center:** A large facility designed to support substantial cloud, AI, or digital infrastructure workloads. **Land Bank:** Land controlled for potential future development where a specific construction timeline may not yet exist. **Powered Shell:** A building with power and core infrastructure installed but without a fully completed customer-specific data hall. **White Space:** The area within a data center where servers, racks, and IT equipment are installed. **Grey Space:** Areas containing supporting mechanical and electrical equipment. **PUE:** Power Usage Effectiveness, calculated as total facility power divided by IT equipment power. ### Project Stages and Dates **Construction Start Date:** The estimated or confirmed date when meaningful physical construction begins. **Construction Completion Date:** The date when major physical construction is substantially complete, which may occur before activation. **Delayed:** A project that remains active but is progressing later than previously expected. **Operational:** A facility that has entered service or is confirmed to be consuming power. **Site Clearing:** Removal of vegetation, structures, or other obstacles before grading and construction. **Site Grading:** Earthmoving activity undertaken to level and prepare a development site. **Vertical Construction:** Construction activity involving the building structure above the foundation. **Commissioning:** Testing and validation of electrical, mechanical, cooling, and operational systems before activation. **Cancelled / Withdrawn:** A project that is no longer expected to proceed under its previously announced plan. ### Power and Grid **Behind-the-Meter (BTM):** Power generation located at or directly connected to the customer site and intended to supply onsite electricity demand. **Balancing Authority (BA):** The organization responsible for balancing electricity supply and demand within a defined region. **Interconnection Queue:** A list of projects requesting permission to connect to the electricity grid. **Interconnection Agreement:** An agreement setting out the technical and commercial requirements for connecting a project to the grid. **ISO / RTO:** Organizations that coordinate wholesale electricity markets and grid operations across defined regions. **Load:** The amount of electricity consumed by a facility or group of facilities. **Megawatt (MW):** A unit of electrical capacity equal to one million watts. **Nameplate Capacity:** The maximum rated output of a power-generation asset under specified conditions. **On-Grid Generation:** Generation connected to and supplying electricity through the broader power grid. **Power Purchase Agreement (PPA):** A contract under which a buyer agrees to purchase electricity from a generation project. **Substation:** Infrastructure used to transform voltage and connect facilities or generation assets to the grid. **Utility:** The company responsible for distributing or supplying electricity within a service territory. ### Power Generation **Battery Energy Storage System (BESS):** A system that stores electricity for later use. **Combined Heat and Power (CHP):** A system that produces electricity and useful thermal energy from the same fuel source. **Distributed Generation:** Smaller-scale generation located near the point of electricity consumption. **Fuel Type:** The primary energy source used by a generation project, such as natural gas, solar, wind, or hydrogen. **Gas Peaker Plant:** A power plant designed to operate during periods of high electricity demand. **Reciprocating Engine:** A combustion engine used to generate electricity, often for flexible or onsite generation. **Transmission Project:** Infrastructure designed to move large amounts of electricity between regions or substations. **Turbine:** Equipment that converts energy from gas, steam, wind, or water into mechanical and electrical power. ### Industrial Developments **Capital Investment:** The disclosed or estimated amount invested in a development project. **Developer:** The company responsible for advancing, financing, or constructing a project. **Early Signal:** An initial indicator of a potential development, such as land acquisition, permits, utility filings, or company disclosures. **Industrial Development:** A major manufacturing, logistics, semiconductor, chemical, energy, or other industrial project. **Project Footprint:** The estimated land or building area occupied by a development. **Site Control:** Evidence that a developer owns, leases, or has an option over the project land. **Special Use Permit:** Local approval allowing a land use that is not automatically permitted under existing zoning. **Zoning / Rezoning:** The local land-use classification or process required to allow a proposed development. ### Aterio Data Fields **Aterio Estimated Power:** Aterio’s estimate of total facility power based on public information, building characteristics, satellite analysis, and proprietary models. **Estimated Date:** A date derived from available evidence when no confirmed public date exists. **Execution Likelihood:** Aterio’s assessment of the probability that a project will proceed based on available development signals. **Published Power:** Power capacity publicly disclosed by a developer, utility, regulator, or another source. **Source URL:** A public source supporting a project field, update, or event. **Unique ID:** A persistent Aterio identifier assigned to a building, campus, company, or power project. **Inventory Table:** The latest available record for each project, campus, or building. **Snapshot:** A complete version of the dataset captured at a specific point in time.