Poveda will use NSF CRII award to develop safer, more efficient autonomous transportation systems
Assistant Professor has received the prestigious Career Initiation Initiative award from the National Science Foundation’s Cyber-Physical Systems Program.
Also referred to as the “mini CAREER Award," this selective award supports untenured junior faculty who have obtained their first academic position after earning their PhD. The award will help fund Poveda’s research and support his graduate students for the next two years.
Poveda won the award for his research titled “High-Performance Adaptive Hybrid Feedback Algorithms for Real-Time Optimization and Learning in Networked Transportation Systems.” It will allow his group to develop new families of adaptive algorithms that better respond to real-time challenges on roadways.
He noted that recent advances in sensing, computation and communication have led to the deployment of new automated technologies on roadways – things like traffic light control and dynamic tolling systems that automatically adjust to changes in traffic conditions.
However, Poveda said, the algorithms these systems use still don’t react well to unexpected “disturbances” like demand fluctuations or traffic accidents. The systems may be relying on inaccurate mathematical models that ignore complex dynamic behaviors, or they may be limited by subsystems that struggle to communicate with one another.
“We still lack the knowledge to deliberately and systematically design robust data-driven feedback algorithms to solve real-time optimization problems in networked transportation systems that exhibit complex and non-smooth dynamics,” Poveda said.
Poveda’s group will start by focusing on the design of adaptive pricing algorithms for the real-time optimization of congestion in traffic roads – a problem that requires tools from areas such as adaptive control, game theory and model-free optimization.
After this, they will focus on the development of advanced data-driven algorithms for traffic lights that are able to autonomously find optimal operating points in real time without any external input from humans. Both technologies are predicted to have a disruptive impact on the performance of the next generation of automated transportation systems.
Poveda was appointed an assistant professor in the Department of Electrical, Computer and Energy Engineering in 2019. He received his PhD in electrical and computer engineering fromthe University of California at Santa Barbara in 2018. Before joining CU Boulder, he also spent time as a postdoctoral fellow at Harvard University.
Poveda is the founding director of the Intelligent Cyber-Physical Systems Laboratory at CU Boulder, which focuses on providing safe autonomy to cyber-physical systems by designing adaptive control and optimization algorithms that combine modern data-driven tools as well as nonlinear control theory for hybrid dynamical systems.