Introduction
The Environmental Protection Agency (EPA) can harness AI to enhance its mission of protecting human health and the environment. By employing innovative applications such as virtual assistants, predictive models, and machine learning techniques, the EPA can improve its ability to manage records, enforce regulations, and assess chemical risks. These AI-driven initiatives have the potential to streamline operations and provide deeper insights and more effective responses to environmental challenges. According to PWC UK, “Research by PwC UK, commissioned by Microsoft, models the economic impact of AI’s application to manage the environment, across four sectors – agriculture, water, energy and transport. It estimates that using AI for environmental applications could contribute up to $5.2 trillion USD to the global economy in 2030, a 4.4% increase relative to business as usual.” The following use cases illustrate how the EPA can utilize AI to advance its regulatory and protective functions, ensuring a healthier and safer environment for all.
Use Cases
-
1. Aidan Chat-bot
Aidan, the virtual assistant developed by the Financial Services Administration (FSA), employs natural language processing to effectively respond to common inquiries regarding financial aid. Over the past two years, Aidan has successfully engaged with more than 2.6 million unique customers, generating over 11 million messages. This high level of interaction demonstrates Aidan’s significant role in providing accessible information about federal aid, thereby enhancing customer service and support.
-
2. Use of random forest model to predict exposure pathways
This study focuses on assessing the potential risks to human health posed by various chemicals by utilizing a random forest model to predict exposure pathways. By analyzing chemical structures and physicochemical properties, the model estimates the likelihood of chemicals being associated with different exposure routes, such as consumer products and dietary sources. The research produced a consensus meta-model that effectively integrates predictions from multiple exposure models, yielding valuable insights into chemical intake rates and identifying specific chemicals that may pose significant health risks.
-
3. Records Categorization
The records management technology team is implementing machine learning techniques to enhance the categorization of records by predicting appropriate retention schedules. This innovative approach will be integrated into a records management application, streamlining the process for users as they submit new records and ensuring compliance with retention policies.
-
4. Enforcement Targeting
In collaboration with the University of Chicago, the EPA’s Office of Compliance has developed a proof-of-concept aimed at enhancing the enforcement of environmental regulations through more effective facility inspections. The implementation of predictive analytics has led to a remarkable 47% increase in the identification of violations related to the Resource Conservation and Recovery Act, thereby strengthening regulatory compliance and environmental protection efforts.