While interviewing for faculty positions I was repeatedly asked how the tool I often use for generating safety filters for controlling robots safely (Hamilton-Jacobi reachability analysis) compares to the very popular control barrier function (CBF) method. I then spent the first couple of years as a faculty member exploring the answer to this along with my students, supported by an ONR YIP.  In this talk, I will discuss:

• The relationship between these two methods
• How we can use reachability analysis to refine CBF approximations to provide safety guarantees
• How we can generalize these safety filters for hard-to-model dynamics (e.g. interaction behavior among pedestrians). This part of the talk is joint with Professor Sicun (Sean) Gao, and recently won the Robocup best paper award at IROS 2023.

Bio: 

Dr. Sylvia Herbert joined the UCSD MAE department as an Assistant Professor after graduating with her PhD from UC Berkeley in 2020.  She now has a research group of 7 PhD students along with several MS and undergraduate students. These students work on a range of projects, including:

- Connections between Hamilton-Jacobi reachability and CBFs (this talk) [1, 2, 3, 4]

- Extensions to safe stabilizing control via control Lyapunov functions (CLFs) [1, 2]

- High-dimensional reachability analysis for very efficient generation of safe control policies [1, 2, 3, 4]
- Reasoning about safety when you are forced to interact with an uncertain stochastic environment (e.g. pushing around debris to reach a goal, surgical robotics) [1, 2, 3]
- Safe reinforcement learning [1]

For more on these topics please see our group website. For questions/comments, Professor Herbert can be reached at sher...@ucsd.edu