KEYNOTE LECTURE

Michele Magno

ETH Zurich, Switzerland

Modern vehicles increasingly rely on a large array of sensors to perceive the environment and support advanced driver assistance and autonomous driving functions. Radar, cameras, inertial sensors, and emerging sensing modalities generate vast amounts of raw data that must be processed reliably, accurately, and in real time. Transforming these signals into actionable intelligence poses significant challenges in terms of latency, energy efficiency, and system scalability. This talk presents recent advances in efficient Edge AI architectures that bridge the gap between sensing and intelligent decision-making directly at the sensor node. By tightly integrating sensing hardware, signal processing, and machine learning, these architectures enable real-time perception while significantly reducing data movement and computational overhead. The presentation will discuss hardware–software co-design strategies for automotive sensing, including energy-efficient embedded AI accelerators, near-sensor processing techniques, and optimized neural network models tailored for radar and multimodal sensing systems. Particular attention will be given to real-time signal interpretation from radar and other sensing technologies, highlighting how edge intelligence can enhance object detection, localization, and situational awareness while operating under strict power and latency constraints. Through selected application examples, the talk will illustrate how edge-native AI pipelines can transform raw sensor signals into reliable high-level information, enabling safer and more responsive automotive systems. The discussion will also highlight future research directions in scalable sensing architectures, distributed intelligence, and next-generation automotive metrology frameworks.

Michele Magno (Fellow Member, IEEE) He is Head of the D-ITET Center for Project-Based Learning Center and Edge AI and Intelligent Sensing Lab. He is also Fellow at IT:U Austria, Michele Magno contributes to the development of new academic initiatives within the IT:U Smart Space Sensing and Systems Lab (S³ Lab) and the IT:U Satellite Lab. His main research interests include wireless sensor networks, wearable devices, machine learning at the edge, smart sensors and autonomous robots, energy harvesting, low-power management techniques, and extending the lifetime of battery-operated devices. He has collaborated with several universities and research centers including industry cooperation with IBM Research, Sony, STMicroelectronics, Infineon, Ferrari and many others. He has published more than 400 papers in international journals and conferences and has received multiple Best Papers and Best Posters awards.