Hi! I'm Nishant.

I'm a 21st-graderfifth-year PhD student and PhD candidate in Robotics at the University of Michigan, advised by Jean-Baptiste Jeannin. My research concerns formal verification for automotive control, such as collision avoidance systems, and has been supported by a NASA Fellowship. Recently, I've been working on automatic verification for collision avoidance, leveraging geometric intuition. We've published a few papers on the "active corner method", which enables automatic generation of efficiently-checkable safe regions for collision avoidance testing at scale. In the past, I've worked as a research intern at Uber ATG (since acquired by Aurora) and on simulation for automated trucks at Ike (since acquired by Nuro). I'm interested in teaching careers after my PhD and am currently a graduate student instructor for EECS 280, an introductory computer science course that's the largest class at Michigan. This summer, I'll be teaching DSC 10 and DSC 40A at UC San Diego.

I went to Berkeley for undergrad, where I studied electrical engineering & computer science, spent five semesters helping teach a data science course that the Wall Street Journal wrote about, won an Outstanding GSI Award, and researched traffic-mitigating strategies for autonomous vehicles and intelligent infrastructure.

At Cal, I was part of the team that built Flow, an open-source framework inferfacing the microscopic vehicle simulator SUMO (and later Aimsun) with the deep reinforcement library Ray RLlib. Flow enables the design and analysis of traffic scenarios as well as the training of agents using deep RL. Here are some papers I contributed to at Cal. I also like to bake, read, and bike.

Email me if you'd like. Here are my resume (short) and CV (long). Find my Google Scholar here.