About me

I am currently a Postdoctoral Researcher at Microsoft Research, NYC. Previously, I obtained the Ph.D. degree in electrical and systems engineering from the University of Pennsylvania (UPenn). I was advised by Prof. Victor M. Preciado. At UPenn, I am also closely working with Prof. Manfred Morari, Prof. George J. Pappas, and Prof. Nikolai Matni. Before coming to UPenn, I received a B.E. degree in Automation from Zhejiang University, China, in 2017.

You can find my CV here and contact me at shaoruchen at microsoft dot com. If you are interested in my research or want to collaborate, please feel free to reach me!

About my research

My research focuses on developing scalable and efficient verification, optimization, and learning tools for safe control of complex nonlinear and learning-enabled systems with formal guarantees. My long-term research goal is to make AI-enabled autonomous systems work safely and reliably in the real world. To achieve this, I believe in the following roadmap: First, develop scalable and efficient tools to analyze the safety/reliability of learning-enabled systems. Then, use the safety/reliability metrics to orchestrate the design of all learning modules in the autonomy stack until a desirable level of guarantee is satisfied.

Following this route, my research has been focused on

  1. Developing specialized optimization tools for NNs: How can we efficiently certify the robustness of NNs and the safety/stability of a dynamical system with NNs in the loop?

  2. Verification-aided learning of safety certificates and safe policies: How should learning and verification interact to generate NN certificates or policies with formal guarantees?

  3. Safe learning-based control: How to correct an unsafe policy online with complex learning dynamics?

I am broadly interested in machine learning, control, and optimization problems that are related to autonomous system design. If you find my work interesting or have any questions, I am happy to have a discussion.