AI & Fundamentals

Bridging Machine Learning and Control for Safe and Intelligent Robot Autonomy - SiQi Zhou, Assistant Professor, Simon Fraser University

Abstract:

As robots become an integral part of our daily lives, they must continuously learn and adapt to operate safely in unstructured, human-centric environments. While established control techniques provide the theoretical foundation for designing high-performance robot systems, their reliance on accurate dynamics models and well-characterized environments often leads to suboptimal performance or unsafe actions when facing real-world uncertainties. This challenge motivates the integration of machine learning into the traditional robot decision-making stack. In our work, we leverage the expressiveness and reasoning capabilities of learned models to enhance robot systems while utilizing expert knowledge from control to safely and effectively integrate these models into embodied systems. In this talk, I will illustrate this concept through three interconnected directions: (1) safe learning-based control that leverages learning methods such as neural networks in combination with control theory to compensate for uncertainties and enable agile robot performance, (2) multi-agent coordination that utilizes distributed control frameworks as a safety filter for reliably deploying high-level task plans generated by large-scale learning models, and (3) perception-based safe decision-making that tightly couples metric-semantic understanding of the environment with control, allowing robots to reason contextually and act safely with “common sense.” These approaches are demonstrated through real-world experiments on various robot platforms, including quadrotors, manipulators, and mobile manipulators. I will conclude this talk with an overview of our broader survey and benchmarking efforts to advance safe real-world robot autonomy, along with an outlook on future directions rooted in principled integration of learning and control for designing robot systems capable of making safe and context-aware decisions in human-centric environments.

Bio:

SiQi Zhou is an Assistant Professor in the School of Computing Science at Simon Fraser University. Her research lies at the intersection of robotics, machine learning, and systems control. By integrating learning techniques with control theory, she develops principled approaches that enable robots to safely and efficiently perform versatile tasks in human-centric environments. SiQi received her Ph.D. and B.A.Sc. degrees in Engineering Science from the University of Toronto in 2022 and 2016, respectively. Before joining SFU, she was a Senior Scientist at the Technical University of Munich and a Postdoctoral Researcher at the Vector Institute. SiQi was selected as one of the MIT Rising Stars in Aerospace (2021), an RSS Pioneer (2022), and a recipient of the EU Marie Skłodowska-Curie Actions (MSCA) Fellowship (2024).

Zoom Link