Towards Data Driven, Inexpensive and Reusable Sensors for Water Contamination Detection

FALL 2018 RESEARCH INCUBATION AWARDEE

PI: Renato Mancuso, Assistant Professor, Computer Science
Co-PI: Xi Ling, Assistant Professor, Chemistry

What is the Challenge?
Drinking water contamination is, sadly, a surprisingly widespread issue. According to a recently published report [1], nearly 77 million people in the U.S. — about a quarter of the total population — in 2015 alone were served by water delivery systems reported in violation of the safety thresholds established by the Safe Drinking Water Act. The problem of poor water quality is hard to address because: (1) water quality testing requires expensive equipment only available in specialized facilities; and (2) testing at the source is often not sufficient because a number of pollutants infiltrate last-mile distribution channels.

What is the Solution?
The solution is to develop a new technology to detect harmful pollutants in drinking water. Three primary design goals for our technology are: (1) low production and maintenance costs; (2) reusability of the solution for repeated measurements over time; and (3) ability to produce water quality assessments in real-time. Because no affordable and reusable sensors exist to detect a number of common contaminants (e.g. lead, copper, nitrates), we use a data-driven approach instead. In a nutshell, our approach consists in using an array of inexpensive, commercially available sensors to collect multiple metrics (e.g. pH, conductivity, …).

What is the Process?
The PIs will implement and refine an embedded device to perform water sampling and analysis. The device will be able to operate with little-to-no human intervention and will be battery-operated. For this endeavor the team will drawn form PI Mancuso’s extensive expertise in embedded and real-time systems. At the same time, PI Ling’s long track record on chemical sensing will be fundamental to navigate the intricacies of data sample preparation and cause-effect correlation in the obtained results.