Plenary Speakers

MAD Games: Multi-Agent Dynamic Game – Lessons from the Limits of Autonomous Racing

Abstract: The critical challenge in deploying autonomous systems is achieving peak performance without compromising safety. Autonomous racing crystallizes this challenge, as it punishes timid policies and demands robust, adaptive strategies in multi-agent settings. Current approaches often fail by either oversimplifying the behavior of other agents or lacking mechanisms for real-time adaptation.
This talk presents research that pushes the boundaries of perception, planning, and control. We will explore how to develop highly competitive agents through:
1. Adversarial Training: Leveraging game theory and distributionally robust online adaptation to create agents that dynamically balance safety and assertiveness.
2. Adaptive Safety: Using conformal prediction, control barrier function and imitation learning we show how multiple imperfect experts train an AI to perform better than any single expert.
3. Safe MPC Frameworks: Implementing an iterative control strategy for nonlinear stochastic systems to handle constrained, real-world uncertainty.
All research is implemented on our F1Tenth/RoboRacer.ai platform—1/10th the size, but 10x the fun. The key takeaway is a deeper understanding of how to build and validate safe autonomous systems for complex, interactive environments.

Biography: Rahul Mangharam is a Professor in the departments of Electrical & Systems Engineering and Computer & Information Science at the University of Pennsylvania, where he directs research on the formal verification and synthesis of safe autonomous systems. His work bridges formal methods, machine learning, and control theory to create provably safe systems for applications including autonomous vehicles, urban air mobility, and life-critical medical devices.
Dr. Mangharam serves as the Penn Director for the $20 million Safety21 National University Transportation Center, a US DOT initiative for safe and efficient mobility. He also directs the Autoware Center of Excellence, an open-source autonomous driving consortium of over 100 industry and academic partners, and is the founder of the F1TENTH Autonomous Racing Community, now active in over 90 universities worldwide.
For his contributions to life-critical systems, he received the Presidential Early Career Award for Scientists and Engineers (PECASE) from President Obama. His work has also been recognized with the NSF CAREER Award, the Intel Early Faculty Career Award, multiple best paper awards from ACM and IEEE, and selection to the National Academy of Engineering’s US Frontiers of Engineering.

Plenary Lecture Ⅱ: November 5(Wed) 13:00-14:00, Premier Ballroom, 2F

Data for Control

Abstract: Computational tools that later evolved into what are now known as semidefinite programs have played a key role in the early development of nonlinear control. These same tools are also powerful to advance data-driven nonlinear control design today. In this presentation, we will examine the basic principles that lead to data-based convex programs for direct controller design and show their effectiveness in solving important control problems.

Biography: Claudio De Persis is a professor with the Engineering and Technology Institute Groningen, University of Groningen, the Netherlands, since 2011. He received the Laurea and PhD degree in electronic and systems engineering in 1996 and 2000, both from the University of Rome “La Sapienza”, Italy. Before joining the University of Groningen, he held postdoctoral and faculty positions at Washington University in St. Louis, Yale University, the University of Rome “La Sapienza” and Twente University. 

Plenary Lecture Ⅲ: November 6(Thu) 10:40-11:40, Premier Ballroom, 2F

Real Haptics for Contact Task by Robot

Abstract: The first question is that why the robot cannot accomplish the simple contact task which the human can do easily? The answer is that the robot does not have the sensation of touch. If we would like to see anything, we need the visual sensation. If we would like to hear anything, we need the audible sensation. In the same manner, if we would like to touch anything, we need the sensation of touch. The real haptics is the technology that digitize, transfer, record, and playback the sensation of touch or the force/tactile sensation (f/t sensation). The real haptics is attained easily by realizing the action-reaction law (Newton’s third law) and the perfect tracking law (synchronization law). But if the two laws hold completely, that leads zero acceleration of both the actuator and the target to touch. This contradiction is solved by introducing both the small delay between these two laws and the employment of the acceleration control. All the necessary process for the real haptics technology is now packed in the custom LSI called AbcCore®.
Secondly how the human motion is caused by the f/t sensation? According to the sensory test, most of our motion is unconscious. The reality is that we decide the motion mission consciously, but the real motion is not conscious. This unconscious motion heavily depends on the f/t sensation. The instantaneous value of the f/t sensation is represented by the value of the force divided by the velocity. By introducing the time window, it is possible to define the intensity and the texture of the f/t sensation. It is interesting that there are four kinds of the receptors under our skin to feel the intensity and the texture of the f/t sensation. By integrating these results, a digitized model of the generation of the motion is constructed such that the motion is a kind of the process of admittance matching by controlling the force and the velocity of the human. The human refers the past experience to cause the motion by the cerebellum. The past experience is a set of the motion data consisting the time history of the force, the velocity, and the f/t sensation. The motion skill is embedded in this motion data. It is possible to design such mechanism in the process of the robot control. In the presentation, the experimental results of the robot motion which the human skill is transferred.

Biography: Kouhei Ohnishi received BE, ME and Ph.D from the University of Tokyo in electrical engineering in 1975, 1977, and 1980 respectively. Since 1980, he has been with Keio University and is a Professor and a Director at the Haptics Research Center in the City of Kawasaki. In 2008 and 2009, he served as a President of the IEEE Industrial Electronics Society. Also he served as a President of the Institute of Electrical Engineers of Japan in 2015 and 2016. His main research includes the motion control, and the real haptics.
He published many papers and books including “Motion Control Systems” Wiley-CRC coauthored with Professor Asif Sabanovic. According to ScholarGPS®, he is one of the highly ranked scholars. He received many awards including the IEEJ paper award (1985), JSPE Award (1996), the IEEE-IES Dr.-Ing Eugine Mittlemann Achievement Award (2004), Fukuzawa Award (2012), Innovation Japan Award (2012), Medal of Honor with Purple Ribbon from His Majesty the Emperor (2016), Fujihara Award (2019) and Hirose Award (2023). He is a Life Fellow of the IEEE and an Honorary Member of the IEEJ.