Skip to main content Start main content

The Right Stuff: Representing Safety for Widespread Autonomy


Image for Event - 30 Jun Seminar
  • Date

    30 Jun 2022

  • Organiser

    Department of Aeronautical and Aviation Engineering

  • Time

    10:00 - 11:00

  • Venue



General Office


Meeting ID: 969 9117 5359 | Passcode: 513687



Autonomous systems, including self-driving cars, delivery drones, and smart infrastructure, have incredible potential to aid people by taking on difficult tasks and augmenting our capabilities. However, it will be difficult to trust such systems in widespread deployment without knowing when they are safe. Compounding the challenge, safety can often be expressed theoretically yet suffer an imperfect translation into numerical representations. So, Dr Kousik’s research focuses on this gap: what representations of safety enable one to implement the theory in a sound way for real-world deployment? This talk will be focusing on safety as collision avoidance for autonomous robot motion planning. Reachability-based Trajectory Design (RTD), a framework that pairs theory with sound numerical implementations will be presented. RTD’s foundation in theory makes it applicable to a wide variety of systems, including self-driving cars, quadrotor drones, and manipulator arms. In practice, over thousands of simulations and dozens of hardware trials, RTD has resulted in no collisions while outperforming other methods, establishing a new state of the art. Excitingly, I find that guaranteeing safety as a baseline can improve performance by allowing one to tune task-related parameters worry-free. Dr Kousik’s future work reaches out from this paradigm to enable autonomous systems to learn and adapt their own notions of safety, with the long-term goal of adaptable deployment in arbitrary settings.



Dr Shreyas Kousik is a postdoctoral scholar in the Department of Aeronautics and Astronautics at Stanford University. He completed his BSc degree in Mechanical Engineering from the Georgia Institute of Technology in 2014 and his MSc and PhD degrees in Mechanical Engineering from the University of Michigan – Ann Arbor in 2020. Dr Kousik’s research focuses on safety in autonomy by seeking geometric and numerical representations to make safe decisions quickly. His PhD work focused on using reachability analysis as a tool to generate such representations for robot motion planning. Currently, he is focusing on reachability applications in state estimation, navigation, and perception. His work has been recognized by the ASME DSCC Best Student Paper award in 2019, the Rackham Merit Fellowship, the Robert J. Beyster Computational Innovation Fellowship, and an honourable mention for the NSF Graduate Research Fellowship.

Your browser is not the latest version. If you continue to browse our website, Some pages may not function properly.

You are recommended to upgrade to a newer version or switch to a different browser. A list of the web browsers that we support can be found here