Autonomy under Uncertainty: Learning for Safety and Coordination
Event details
| Date | 15.02.2021 |
| Hour | 15:00 › 16:00 |
| Speaker | Prof. Maryam Kamgarpour, Electrical and Computer Engineering, The University of British Columbia |
| Location | Online |
| Category | Conferences - Seminars |
Abstract: We began using robots to help us with simple and repetitive tasks. Now, we are automating increasingly safety-critical and complex systems: power grids, transportation systems, search and rescue missions, financial markets, etc.
I present two of my research directions addressing some of the challenges these autonomous systems exhibit, namely, increased uncertainties due to imperfect models or due to the large-scale interactions of the autonomous systems.
First, I discuss control design for ensuring system safety. I focus on algorithms with provable guarantees in the face of increased uncertainties on the one hand, and the availability of data on the other. Second, I discuss coordinating decision-making of a large number of interacting autonomous systems. Here, I address learning-based approaches, so as to account for availability of local data. I conclude with reflecting on open directions.
Bio: Maryam Kamgarpour holds a Doctor of Philosophy in Engineering from the University of California, Berkeley and a Bachelor of Applied Science from University of Waterloo, Canada. Her research is on safe decision-making and control under uncertainty, game theory and mechanism design, mixed integer and stochastic optimization and control. Her theoretical research is motivated by control challenges arising in intelligent transportation networks, robotics, power grid systems, financial markets and healthcare. She is the recipient of NASA High Potential Individual Award, NASA Excellence in Publication Award, the European Union (ERC) Starting Grant and NSERC Discovery Accelerator Grant.
I present two of my research directions addressing some of the challenges these autonomous systems exhibit, namely, increased uncertainties due to imperfect models or due to the large-scale interactions of the autonomous systems.
First, I discuss control design for ensuring system safety. I focus on algorithms with provable guarantees in the face of increased uncertainties on the one hand, and the availability of data on the other. Second, I discuss coordinating decision-making of a large number of interacting autonomous systems. Here, I address learning-based approaches, so as to account for availability of local data. I conclude with reflecting on open directions.
Bio: Maryam Kamgarpour holds a Doctor of Philosophy in Engineering from the University of California, Berkeley and a Bachelor of Applied Science from University of Waterloo, Canada. Her research is on safe decision-making and control under uncertainty, game theory and mechanism design, mixed integer and stochastic optimization and control. Her theoretical research is motivated by control challenges arising in intelligent transportation networks, robotics, power grid systems, financial markets and healthcare. She is the recipient of NASA High Potential Individual Award, NASA Excellence in Publication Award, the European Union (ERC) Starting Grant and NSERC Discovery Accelerator Grant.
Practical information
- General public
- Free