By Akshay Rajhans

Multi-Paradigm Modeling for Design and Operation of Intelligent Cyber-Physical Systems

Abstract

Advances in computation and communication technologies have fueled the recent rise in intelligent cyber-physical systems. Such systems equipped with interconnected computational elements for sensing and controlling their physical environment are now increasingly getting embedded with some level of artificial intelligence in them. With these advances in autonomy, such systems will be able to cooperate and collaborate with humans and other artificial agents, potentially in situations they have never encountered before. Ensuring not just their correct design but their correct operation in presence of all unforeseen operating scenarios is now a research grand-challenge. Computational modeling for design and analysis has the potential of providing the necessary rigor, yet because this domain brings together traditionally disparate engineering disciplines – computer science, control theory, communication, signal processing, and now machine learning, to name a few – the respective mathematical abstractions and models of computation transcend the traditional boundaries. This talk presents a perspective based on a decade spent working on multi-paradigm modeling for CPS, first as an academic formally reasoning about the heterogeneity, and then as an industrial software tool developer building bridges between modeling formalisms. Challenges and opportunities ahead in terms of theory, practice, and bridging the gap between the two will also be outlined.

Bio

Dr. Akshay Rajhans is a Principal Research Scientist in the Advanced Research & Technology Office at MathWorks. His work and research interests lie in multi-formalism model-based design of cyber-physical systems (CPS). He organizes and chairs research conferences including the MathWorks Research Summits, the CPS tracks at Winter Simulation Conference and Spring Simulation Conference, and the Workshop on Monitoring and Testing for CPS (MT-CPS) at CPS-IoT Week 2019. He has been an invited speaker and a panelist at various research gatherings for over a decade, and has served on Editorial and Industry Advisory Boards of various universities, conferences, and professional organizations, as well as on Ph.D. Thesis Committees and Technical Committees. As a CPS practitioner, he previously worked on electronic control for diesel engine applications at Cummins and on non-intrusive load monitoring at Bosch Research. He has a Ph.D. from Carnegie Mellon University and an M.S. from University of Pennsylvania. He is a member of IEEE and ACM.