RICHLAND, Wash.-For the first time in the over 50-year history of the National Environmental Policy Act, federal regulators now have vital information at their fingertips in a single database, thanks to an ambitious project called PermitAI, developed by the Department of Energy's Pacific Northwest National Laboratory.
Previously, NEPA review data resided in siloed federal computer systems, often in document formats that cannot be easily processed by modern data-crunching methods, such as artificial intelligence.
The research team at PNNL solved the silo problem with clever use of an advanced, custom AI algorithm trained to automatically recognize and categorize data. Once processed, the machine-readable dataset, known as NEPA Text Corpus (NEPATEC), makes the labor-intensive searches much simpler.
The PermitAI team has now released its AI-ready, open-access NEPATEC 2.0 database to the public.
"We are improving the quality and accessibility of environmental data using a custom AI toolset. This has never been done before. We are doing the hard work, so the users don't have to," said Sameera Horawalavithana, a PNNL data scientist and co-lead of the PermitAI team. "Interagency coordination has proven to be a challenge historically in federal environmental reviews. Delays can be extremely costly in time and money. With these advancements, PermitAI is making it easier to compare, combine and analyze information accurately, cutting down on delays and streamlining review processing."
In accordance with America's AI Action Plan and the Council on Environmental Quality's directive for federal agencies to adopt uniform data standards, NEPATEC 2.0 has expanded to include data from multiple federal agencies involved in making environmental permitting decisions. These data include environmental impact statements, environmental assessments, categorical exclusions and other key NEPA planning documents that inform decision-makers.
"The release of NEPATEC 2.0 marks a key milestone in improving environmental review," said Neelesh Nerurkar, Director of Infrastructure Policy at DOE's Office of Policy. "By making the permitting records usable at scale, it equips project developers and agencies with the data needed to reduce review timelines for crucial energy investments without sacrificing quality. PermitAI enables us to build the infrastructure we need for the future faster, smarter and more efficiently."
Breaking down silos
PermitAI began as a pilot project sponsored by the Department's Office of Policy to centralize NEPA decision data for the DOE. Driven by the recent federal priority to accelerate and improve environmental reviews, the PermitAI team has expanded its focus.
Now, four federal agencies, including the Department of the Interior, the U.S. Army Corps of Engineers, the National Institute of Standards and Technology, and the Department of Agriculture, have granted PermitAI access to their data, encompassing around 120,000 documents from 60,000 projects authored by more than 60 different agencies. And the number of participating agencies is expected to rise. Read the full report about the dataset construction.
The federal government is seeking to modernize the electric grid infrastructure, build data centers and site new semiconductor manufacturing facilities, among other priorities. All of these efforts require NEPA reviews. Federal agency staff will now have access to key data across multiple agencies in minutes instead of months, speeding up their ability to make informed decisions.
NEPA reviews provide the public with an opportunity to better understand the environmental impacts of proposed projects in their communities. But the process often involves multiple agencies working on different aspects of land use or other proposals, which can be complex and challenging to coordinate. PermitAI aims to simplify this coordination, creating a more efficient space for collaboration.
Why modern data standards matter
If you've ever tried to understand a street address in a foreign country, you quickly grasp that information can be confusing when it doesn't conform to the format you expect. Absent agreed-upon rules and guidelines for an address's structure, trying to compare across nations quickly becomes a jumbled mess. That's why data standards matter.
NEPATEC 2.0 uses a NEPA- and permitting-specific data standard issued by the Council on Environmental Quality to streamline and standardize data and categorize it through "metadata" that provides key information such as the relationships between data points. This coded metadata makes information easier to understand and reuse.
"We are not just collecting the data, we are doing a lot of data enrichment that will have utility to federal agencies and the public," said Sai Munikoti, a PNNL data scientist whose work involves development and application of specialized large language models. Munikoti is a PermitAI co-lead who works closely with federal agencies implementing data standards for environmental permitting.
As part of PNNL's strategic public-private partnership model shepherded by the Center for Continuum Computing, the PermitAI team also collaborated with Google to implement data enrichment and storage in the Google Cloud Platform, making data sharing easier.
"Our collaboration with PNNL is powering this critical project with cutting-edge technology and AI capabilities that are helping to transform environmental permitting with accessible and secure data," said Cameron Groves, Director, Google Public Sector.
A responsive and interactive search tool
In addition to the public database, the PermitAI team is currently beta testing SearchNEPA, an interactive AI-driven toolkit with an interface designed for federal NEPA reviewers. More than 200 federal evaluators are currently putting the AI component to the test. SearchNEPA provides a plain language interface for NEPATEC's more than 120,000 NEPA documents and decisions spanning 50-plus years of data collection, with more coming on-line regularly as agencies comply with the federal mandate to adopt uniform standards.
Future applications now under development include EngageNEPA and CommentNEPA. The ongoing research is funded by the DOE Office of Policy.
A recent workshop hosted by PNNL showcased some of the early uses of PermitAI.
To stay up to date on PermitAI and important news related to AI in federal permitting, subscribe to the PermitAI newsletter.
