Research Project Title:
“The Car Can Explain” — Propagators for Autonomous Vehicles
abstract:As usage of autonomous vehicles becomes more prominent we must be thinking about potential vehicle accidents and errors. The Car Can Explain project is a multi-year Toyota Research Initiative/Computer Science and Artificial Intelligence Laboratory (CSAIL) project that aims to explain vehicle actions and develop a symbolic, explanatory diagnostic system for the vehicle using constraint propagation qualitative and semi-qualitative reasoning. The original models use measured data to initialize the model parameters and then propagate symbolic values to produce concise, understandable explanations of vehicle actions. To the human operator, this will be perceived as a causal chain of reasoning. For my SuperUROP project, I will focus on contributing to the fault detection code as well as proposing fair and practical fault assignment protocols.
“Epitomizing MIT’s motto of mens et manus (“mind and hand”), this SuperUROP project provides me with a great opportunity to apply my academic knowledge and skills in solving real-world problems. I look forward to integrating my double-major background in computer science and mechanical engineering to advance the development of autonomous vehicles.”