报告时间：2018年4月23日周一15:00

报告地点：湖南大学信息科学与工程学院106教室

报告人简介：

    常万里博士现任教于新加坡理工大学，近期将前往英国约克大学任教。他于2017年9月从慕尼黑工大电

子与计算机工程学院获得博士学位以及最杰出博士论文奖（Summa Cum Laude）。第一指导教授为Samarjit

Chakraborty（TUM），第二指导教授为Anuradha Annaswamy（MIT）。

    他在国际顶尖学术期刊和会议发表多篇论文，包括IEEE旗舰期刊Proceedings of the IEEE（IF=9.23

7），曾获嵌入式系统顶级会议CODES+ISSS最佳论文提名奖和RTCSA最佳论文奖。常博士的论文共同作者和合

著者包括多名IEEE Fellow, ACM Fellow，美国国家工程院院士，欧洲科学院院士，德国科学工程院院

士。    

    项目合作公司包括Bosch，Daimler，Audi，General Motor，Google，新加坡港务集团，新加坡陈笃

生医院等。他目前主持一项新加坡国家级科研项目（第二级，约相当于250万人民币），新任两本SCI期刊副

主编，Microprocessors and Microsystems（CCF C），Journal of Circuits, Systems and Compute

rs，以及Journal of Systems Architecture（CCF B）的客座编辑。常博士是设计自动化顶会DAC 2018的

分会主席（副主席为普度大学副教授Vijay Raghunathan），以及RTSS 2018（CCF A）的技术委员会成员。

报告内容：

Abstract: Conventionally, there is a separate design paradigm for cyber-physical systems (CPS). Control algorithms interacting with the physical processes are designed without considering details of the computing implementation architecture. There is a significant inefficiency of resource utilization and large safety margins are added. In addition, metrics that are relatively simple to analyze are targeted, such as stability and quadratic cost. The above design paradigm is only possible for simple applications. As the intelligent era is emerging, the applications and functions of CPS are increasing in both complexity and number. Examples are autonomous vehicles and smart robots. On one hand, more trusted and certifiable CPS are required, and on the other hand, it is necessary to maximize the control performance with the given resources, or minimize the resources for the constraint on the control performance. This motivates trusted and resource-aware CPS design. While communication-aware control systems design has been extensively investigated by the NCS community, other implementation resources, including the memory hierarchy, computation, operating system, energy, etc. also need to be considered. This talk presents a pioneer work on memory-aware CPS design.