Abstract
Digital twinning is gaining popularity in domains outside traditionally engineered systems, including cyber-physical systems with biological modalities, i.e., cyber-biophysical systems (CBPS). While digital twinning has well-established practices in cyber-physical systems settings, it raises unique challenges in the context of CBPS. In this talk, we discuss the digital twin we have been working on for CBPS in controlled environment agriculture. We present how the plants and their environment can be modeled and how reinforcement learning can help construct the complex underlying simulators. We will end with a discussion about challenges and lessons learned in developing digital twins for CBPS.
Bio
Eugene Syriani is a full professor in Computer Science at the University of Montreal. He received his Ph.D. in Computer Science in 2011 from McGill University. His main research interests fall in software design based on the model-driven engineering approach and simulation-based design of autonomous systems. On the software engineering side, he regularly contributes to techniques for domain-specific modeling, model transformation, collaborative modeling, and the generation of fully customizable modeling environments to improve user experience. On the simulation side, he works on digital twins, discrete event simulation, and co-simulation. He is currently leading several research projects on these topics in Canada and has been leading the development of several publicly available software tools.
