Professor Dezong Zhao from University of Glasgow Delivers an Academic Lecture at GIFT: Efficient, Safe, Trustworthy and Causal Autonomous Systems

On October 20, 2025, Professor Dezong Zhao from the James Watt School of Engineering, University of Glasgow was invited to deliver an academic lecture titled "Efficient, Safe, Trustworthy and Causal Autonomous Systems" at GIFT. The lecture was organized by the Innovation Center of Intelligent Connected Electric Vehicles and attended by over 20 faculty members and students from the center.

Professor Zhao is the recipient of Turing Fellowship, Royal Society Newton Advanced Fellowship, Royal Academy of Engineering/Leverhulme Trust Research Fellowship and UK EPSRC Fellowship. He has long been dedicated to research areas including autonomous vehicles, robotics, causal inference and machine learning. In his presentation, he shared his team's latest advancements in driverless vehicles and autonomous systems, focusing on three core aspects: efficiency, safety and trustworthiness.

In terms of efficiency, Professor Zhao highlighted energy conservation and intelligent optimization for unmanned systems, introducing energy consumption control methods and efficient perception technologies ranging from single vehicles to entire fleets.

Regarding safety and trustworthiness, he explored the modeling and decision-making mechanisms of autonomous systems in complex environments, emphasizing the need for self-correction and adaptive capabilities. He also demonstrated behavior prediction methods based on game theory and graph neural networks, and proposed integrating causal inference to enhance system interpretability and reliability.

Concluding his talk, Professor Zhao shared his team's recent explorations into causal autonomous systems. He highlighted that true intelligence should not only rely on data correlations, but must also understand causal relationships in the world.

During the Q&A session, faculty and students engaged in lively discussions on topics such as simulation environment construction, scenario generation strategies and the integration of causal inference with deep learning.

The lecture is thought-provoking and informative. It showcased the latest advances in autonomous systems and offered valuable academic inspiration for GIFT participants. The event concluded with warm applause.