Ken Foo
EmNet "was formed in 2004 in partnership with the City of South Bend (Indiana) and the University of Notre Dame, for the specific purpose of solving South Bend’s combined sewer overflow problem using advanced control systems theory and civil engineering." The result of this effort was that the combined sewer overflow in the City was reduced by 70%. Additionally, this project cost about a 1/10th of the $100 million it would have cost using traditional engineering methods. Says South Bend's Mayor Pete, “We have the smartest sewers in the world!”
For my ESTEEM capstone thesis project I am applying the same concepts of machine learning and advanced controls from the sewer overflow project to the City's wastewater treatment plant. The objective is to apply machine learning to optimize water treatment to provide the cleanest water while using the least amount of energy. Ironically, machine learning has been applied to Google searches, Netflix suggestions and Facebook feeds, but it has rarely been applied to life’s most important substance: drinking water.
These past few months, as I have focused more on customer interviews, I have learned a tremendous amount about the wastewater industry. Now, I will start profiling each customer segment and begin to tailor a solution for each. Also, I have begun to interview water treatment innovators and startup companies to learn more. I look forward to what I will learn as I develop a financial and marketing strategy and complete my thesis project.