#348 Preparing for Pothole-Filling Robots: Analytics, Routing, and Ethics

In 2018, the Chicago Department of Transportation (CDOT) filled 40,225 potholes around the city. On average, CDOT completed requests 2.5 weeks after they were reported. How can undergraduate teams improve Chicago’s pothole situation on a cold-start 135-hour budget for analytics and mathematical modeling?

Two teams of applied math and computer science students from Illinois Tech’s Math 497 class Problem-Solving Projects in Business, Government and Industry share their progress exploring the Chicago Data Portal’s pothole data and designing placement and routing algorithms to prepare for future self-driving robotic pothole fillers. Civic analytics advocate David Eads will introduce the teams, and machine-learning ethics proponent Vinesh Kannan will wrap things up with a participatory pothole ethics exercise.

P++ Team: Sarbani Bhattacharyya, Adriana Cortes, Anna Kong, and Jelena Mihajlovic-Klaric. Using exploratory data analysis to inform the placement of robot bases and predict intensity of demand, the P++ team will show how to employ the Capacitated Facility Location and Chinese Postman problems from graph theory to locate the bases and route the robots to fill potholes.

Pothole Detectives Team: Luke Logan, Erik Quintanilla, and Ekrem Güzelyel. The Pothole Detectives team will show disparities in pothole-filling wait times across Chicago wards, and propose another robot-routing algorithm based on graph-theoretic decomposition of street maps for prioritized filling of potholes.

Math 497 Advisor: Robert Ellis, Associate Professor of Applied Mathematics, Illinois Tech

RSVP required Braintree now requires all attendees to RSVP beforehand by 12:00 PM (noon). Walk-ins will not be allowed!

ASL This event will have an American Sign Language interpreter.

Food Food and drinks will be provided. We encourage attendees to bring their own water bottles to reduce waste.