Water Resources Research Act Program

Details for Project ID LA_2025_Bappy

Interpretable Predictive Modeling for Dissolved Oxygen Monitoring and Proactive Aeration Control in Rural Wastewater Treatment Systems

Institute: Louisiana
Year Established: 2025 Start Date: 2026-01-01 End Date: 2026-12-31
Total Federal Funds: $15,550 Total Non-Federal Funds: $15,551

Principal Investigators: Mahathir Bappy and Michael Hayes

Project Summary: Louisiana's rural wastewater treatment systems are important for protecting community health and local waterways. However, these systems often struggle due to limited budgets, changing weather conditions, and aging equipment. One critical challenge is managing dissolved oxygen (DO), which helps remove pollutants during treatment. Poor DO control can lead to inefficient treatment and unnecessary energy use. To solve this problem, our study introduces an easy-to-understand artificial intelligence (AI) tool called TimeGPT to help predict future DO levels. The tool analyzes past measurements along with other water quality indicators, such as temperature, pH, turbidity, nitrate, and ammonium, to anticipate when DO might drop or fluctuate. To ensure the tool is trustworthy and explainable, we use SHAP value visual explanations that show which factors influence each prediction. This helps plant staff understand how the model works and why specific forecasts are made. By supporting smarter aeration decisions, this AI tool can help reduce energy costs, improve treatment reliability, and support cleaner, more sustainable wastewater management for Louisiana's rural communities.