Title: Towards flexible yet safe autonomous systems – blending machine learning with predictive control.
Time: Tuesday, 1st June, 16:00 CET (10:00 EST).
There is a steadily increasing need for autonomous systems that can operate independently, adapt and learn to tackle changing conditions. Applications span from medical robotics, autonomous cars, industrial production systems to chemical industries and food processing. Control and decision-making often are at the core of the underlying technologies. Fusing control approaches with machine learning methods are widely seen as fundamental to achieve autonomy. Despite significant advances, using machine learning approaches in control and automation is still in its infancy. One of the main reasons for this is the need for the dependable, explainable, and safe operation of autonomous systems. Blending predictive control approaches with machine learning approaches to achieve safety and reliability is promising. It allows to adapt to changes and to use data-driven and hybrid models. After a motivation, we outline strategies to blend predictive control and learning approaches, focusing on performance and stability guarantees applied to robotics, autonomous driving, and chemical processes.
Prof Rolf Findeisen received an MSc degree from the University of Wisconsin–Madison, Madison, WI, USA, a Diploma degree from the University in Stuttgart in Engineering Cybernetics, and a Dr. degree from the University of Stuttgart, Stuttgart, Germany, in 2005. He was a Research Assistant with the Automatic Control Laboratory, ETH Zurich, Switzerland, and a Researcher with the Institute for Systems Theory and Automatic Control, University of Stuttgart. Rolf heads the Systems Theory and Automatic Control Laboratory at the Otto-von-Guericke University Magdeburg, Germany. He had research stays and guest professorships at the Massachusetts Institute of Technology Cambridge, EPF Lausanne, the University of California at Santa Barbara, Imperial College London, NTNU Trondheim, Norway. Rolf’s current research activities focus on the control and monitoring of autonomous and cyber-physical systems, predictive control, and the fusion of machine learning and control with guarantees. Fields of applications span from mechatronics, aerospace systems to chemical and biotechnological processes, robotics, energy systems, and systems medicine.
The AI Colloquium – Invited Speech
Participation is FREE and OPEN to all.
Prior REGISTRATION is required.