Safety Critical
Safety-critical systems research focuses on designing and verifying systems where failures can have catastrophic consequences, with a current emphasis on robotics, autonomous vehicles, and AI. Key research areas involve developing robust algorithms for real-time safety monitoring and recovery (e.g., using Q-networks and control barrier functions), diagnosing and mitigating distribution shifts in machine learning models (e.g., via martingales), and ensuring fairness and efficiency in multi-agent systems. This work is crucial for deploying reliable and trustworthy autonomous systems across various domains, improving safety and enabling wider adoption of advanced technologies.
18papers
Papers
October 28, 2024
September 16, 2024
February 7, 2024
January 26, 2023
September 26, 2022
Digital Twin in Safety-Critical Robotics Applications: Opportunities and Challenges
Sabur Baidya, Sumit K. Das, Mohammad Helal Uddin, Chase Kosek, Chris SummersEdge-assisted Collaborative Digital Twin for Safety-Critical Robotics in Industrial IoT
Sumit K. Das, Mohammad Helal Uddin, Sabur BaidyaRisk-Aware Model Predictive Path Integral Control Using Conditional Value-at-Risk
Ji Yin, Zhiyuan Zhang, Panagiotis Tsiotras