WASP PhD students Alejandro Luque Cerpa and Mengyuan Wang, together with their supervisor Hazem Torfah (Chalmers University of Technology), have received the Best Paper Award in the Learning-Enabled Autonomy Track at HSCC/ICCPS 2026 for their work on safe AI-based autonomy. The paper was co-authored with WASP researcher Devdatt Dubhashi at Chalmers, Emil Carlsson at Sleep Cycle, and Sanjit Seshia, Professor at EECS at the University of California, Berkeley.

The awarded paper, Learning Contextual Runtime Monitors for Safe AI-Based Autonomy, introduces a novel framework designed to strengthen the safe operation of AI-based control systems. The work focuses on the development of context-aware runtime monitors capable of assessing system behaviour in real time and supporting safer decision-making in complex environments.

“Our work addresses the safe use of machine learning-based control in AI-driven autonomy. It introduces a framework for learning context-aware runtime monitors for AI-based control ensembles, combining runtime verification techniques with contextual multi-armed bandit learning,” says Alejandro Luque Cerpa, WASP PhD student at Chalmers University of Technology.

The researchers emphasize the broader importance of their work. “We believe that autonomous systems will play a major role in shaping the future. It is therefore essential to ensure that they are safe. No one would trust or use an autonomous car without confidence in its reliability. Every step toward making autonomous systems safer has an important impact. Dynamic approaches such as runtime monitoring are key to enabling autonomous systems to adapt safely to complex and changing environments.”

Paper details

Title: Learning Contextual Runtime Monitors for Safe AI-Based Autonomy
Authors: Alejandro Luque Cerpa, Mengyuan Wang, Emil Carlsson, Sanjit A. Seshia, Devdatt Dubhashi, Hazem Torfah