PhD student wins ACM SIGBED's Student Research Competition
Illinois ECE PhD student Hussein Sibai recently won the gold prize at the ACM SIGBED's (Special Interest Group in emBEDded systems) Student Research Competition. This competition happens alongside CPS-IOT Week, which is a major venue in the research community. CPS-IOT week hosts five parallel conferences on Cyber-Physical Systems and Internet-of-Things covering the whole research range from theory to practice. Advised by ECE Professor Sayan Mitra, Sibai developed an algorithm for verifying autonomous systems that exploit the underlying structure of the dynamics and dramatically speed up existing approaches. The work is being applied to autonomous vehicles and urban air mobility systems.
"My poster and presentation in the competition were based on the core contribution in my PhD thesis," said Sibai. "For me, this award shows appreciation to my thesis work by the research community, and it excites me that they see my work useful, the goal of any researcher. It is a beautiful way to finish my PhD and start applying for jobs."
Sibai's research focused on developing efficient methods to verify and validate the safety of cyber-physical systems, such as drones and autonomous vehicles. A typical example would be verifying that a drone following a predefined path in a cluttered environment will not collide with any obstacles, despite sensing and state estimation errors. However, existing verification methods are challenged with the scalability problem to more complex scenarios, and huge efforts are being spent on finding efficient ones.
"My research utilized symmetry, one of the most important properties of many physical systems, to accelerate their safety verification," added Sibai. "I developed an efficient verification algorithm based on symmetry. The algorithm reduces the model of the system before verifying it, then refines the reduced model to reach the right level of reduction that is accurate and easy to verify. I, along with my colleague Yangge Li, developed a software tool based on the algorithm. We achieved an order-of-magnitude speedup in verification time for drones and cars following complex paths in cluttered environments. The algorithm is just the start for symmetry utilization in designing safe systems with many open problems to be solved."
"I would like to thank my advisor, Prof. Sayan Mitra, for his continuous guidance and support throughout my graduate studies and my research that led up to this award, my colleague Yannge Li for his immense help in developing the tool, to the judges and chairs of the competition for their efforts and recognition, to the other competitors that I was pleased to meet and learn from, to ACM, and Microsoft, the sponsor of the competition. "