讲座时间：2018年5月31日下午16:30

讲座地点：学院542 会议室

讲座简介：Cyber-Physical Systems (CPS) must be dependable due to their safety-critical nature. In the meantime, the recent trend to collect a large amount of data from various sensing and communication capabilities has enabled modern embedded software to be more adaptive. Examples include connected and autonomous vehicles, UAVs, and warehouse robots. There are grand challenges to design and operate such systems due to their seemly conflicting requirements: adaptivity needs adequate response to various, often unknown and unpredictable, internal and external situations, while dependability must ensure predictable correctness.

In this talk I present a collection of technical solutions. One is runtime resource management for time-critical CPS software, that dynamically adjust scheduling and software parameters according to system dynamics while guaranteeing real-time reliability (e.g., all timing requirements are predictably satisfied). The other is runtime safety enforcement for CPS through the use of safety guards, which are reactive components synthesized automatically from safety specifications to protect against catastrophic failures. The approach is particularly suitable for adaptive systems that continuously evolve with new data.

讲座人介绍: Haibo Zeng is currently a faculty member at Virginia Tech, USA. He received his Ph.D. in Electrical Engineering and Computer Sciences from University of California at Berkeley, a B.E. and M.E. in Electrical Engineering from Tsinghua University, Beijing, China. He was a senior researcher at General Motors R&D until October 2011, and a faculty member at McGill University,  Canada from November 2011 to August 2014. He has published over 90 peer reviewed journal and conference papers, many at top venues. His work has received four best paper/best student paper awards (including those at RTSS’17 and ECRTS’13) and three other best paper nominations.