Towards Continuous Assurance for Autonomy Software
|
ABSTRACT In this project, we have studied (1) continuous integration (CI) pipelines for safety testing of autonomous systems, (2) possible abstractions for perception modules used in autonomous systems, and (3) the role of nondeterminism in testing and verification. We have setup a CI pipeline for an autonomous car racing competition using Jenkins, Docker, and AWS. From our experiments, we found that nondeterminism, primarily arising from communication delays and physics simulation, can significantly impact the test outcomes. Our literature review confirms similar observations from other studies. Slides found here. |
BIO
Sayan Mitra is a Professor of Electrical and Computer Engineering at the University of Illinois at Urbana-Champaign. He is the Associate Director of the Center for Autonomy at Illinois. His textbook ”Verifying Cyber-Physical Systems: A Path to Safe Autonomy” was published by MIT Press in 2021. His group has produced several algorithms and software ools for synthesis and formal verification. His PhD advisee Chuchu Fan’s thesis won the ACM Doctoral Dissertation Award in 2020, which is given to the best Computer Science PhD of the year. Sayan holds a PhD from MIT. His work has been recognized by the National Science Foundation's CAREER Award, AFOSR Young Investigator Award, IEEE-HKN C. Holmes MacDonald Outstanding Teaching Award, a RiSE Fellowship, a Siebel Fellowship, UIUC Dean’s Award for Excellence in Research, and several best paper awards.