Practical deployment of a fully autonomous vehicle has been a goal of many machine learning groups. Recent advancements coupled with availability of mobile location services make today a fitting time to deploy driverless cars to perform taxi/ridesharing services. This project aims to deploy a group of electric cars that can be summoned to users’ locations across MIT via a mobile application or the web, and then automatically take them to their destination. The data-collection phase will also create maps of MIT that will be useful in training future autonomous systems. Success of this project would serve as a proof of concept of automated vehicles operating in an urban environment alongside humans without a centralized system.

I grew up with an interest in the automotive and transportation industries so this project resonated with me. I’ve also been interested in electric power being used for widespread personal transport, and hopefully this project will gain me some more exposure to that as well. Before I started this SuperUROP, I worked in a UROP with the MVL to make a head-tracking HUD system for use in trains.