Case Study: Anacostia Riverkeeper achieves timely E. coli predictions for improved water-quality monitoring with DataRobot

Anacostia Riverkeeper Uses DataRobot to Predict Water Quality in the Anacostia River

Anacostia Riverkeeper, a Washington, DC nonprofit dedicated to protecting and restoring the Anacostia River, faced a challenge: lab tests for E. coli take days to return results while water quality can change rapidly with weather, leaving the public without timely information. Through DataRobot’s AI for Good program, Anacostia Riverkeeper partnered with DataRobot to develop predictive models to flag when E. coli levels exceed safe thresholds.

Using DataRobot, the team trained and deployed dozens of binary classification models fed by real‑time USGS sensor data from 28 sites (discharge, gauge height, temperature, etc.), with features aggregated over 12–24 hour windows. A pipeline pulls sensor feeds, sends them to DataRobot for scoring, stores results in a database, and visualizes forecasts in Tableau—enabling multiple daily predictions that shorten the delay between sampling and actionable information. DataRobot’s solution delivered operational, shareable forecasts that enhanced river monitoring and has been shared with stakeholders across the Chesapeake Bay region.

Anacostia Riverkeeper

Olivia Anderson

Former Project Coordinator