Speaker: Yanbing Mao, Engineering Technology, Wayne State University

Time: Wednesday, February 25, from 2:30 pm to 3:30 pm

Location for in-person participants: 1146 FAB 

Zoom link for online audience: https://wayne-edu.zoom.us/j/92845590121?pwd=CpRA5Wa5gzSMn2xiVkR2abD83O5nrH.1

Title: Edge Brain: Cognitively-Grounded Runtime Learning on Edge

Abstract: AI offers unprecedented opportunities to revolutionize robotics by achieving human-level intelligent automation. However, this prospect faces several critical challenges. First, prevalent AI strategies involve training a policy in a simulator and then deploying it to a real-world system. But the physical world is usually non-stationary, hard-to-predict, hard-to-simulate, and uncontrollable, creating significant Sim2Real and domain gaps. Second, many robotic systems, such as legged robots and autonomous vehicles, are safety- and time-critical. A minor AI fault can lead to catastrophic failures. Third, learning often encounters imbalance and misalignment due to corner cases and random sampling. Cognitive studies have revealed that humans possess inherent cognitive capabilities that naturally address these challenges. Building on advances in cognitive science, we developed the Edge Brain. Equipped with Edge brain, safety-critical robotics is able to perform human-like runtime learning in the complex physical world. Additionally, Edge Brain eliminates the need for GPS, Wi-Fi, and wireless communication -- everything (perception, learning, and control) is on edge. This enables broader deployment of safety- and time-critical applications in demanding physical environments, such as deep forests and post-disaster areas with obstructed communications. In this talk, I will present the development of Edge Brain in addressing these challenges, and its demonstrations by quadruped robots.

Bio: Dr. Mao is an Assistant Professor with the Engineering Technology Division at Wayne State University, MI, USA. He received his Ph.D. degree in Electrical and Computer Engineering from the State University of New York at Binghamton in 2019. From 2019 to 2022, he was a Postdoctoral Researcher at the University of Illinois Urbana–Champaign. His current research focuses on robotics and AI, with a focus on safety-critical applications. His research has been funded by NSF and NVIDIA.