Control theory is the mathematics behind feedback systems that ensure your room is at the exact temperature you require or your car cruises at the exact speed you desire. Master control theory and robots will do whatever you tell them.
At Mason, we develop and utilize advanced control-theoretic tools to design, build, and deploy large teams of heterogeneous robots working together to complete various tasks. Our areas of expertise include:
Scheduling and Control Co-design for Cyber Physical Human Systems
Cyber Physical Human Systems (CPHSs) represent an emerging multi-disciplinary field that has seen fast recent growth with the advancement of autonomy and networked communication. Our research in CPHS reaches unexplored territory in the intersection of control theory and real-time systems theory. The research goal is to maximize the benefit of the advancement in information technology, control theory, real-time scheduling and computational science to enable more social and economic benefits for future human society. The general applications include Cyber-Physical Systems (CPS), the Internet of Things (IoT), intelligent traffic management and human robot interaction. Principal investigator: Ningshi Yao
Human and Robot Interaction
With increasing penetration of robots in industry and everyday life, cooperation between humans and robots is quickly becoming unavoidable. It is extremely important that robot interact with humans safely and naturally, and to this end, the study of human robot interaction (HRI) has enjoyed recent research interests. Research in this domain require a good balance between model-based methods and learning-based methods, and between fully autonomy and human guidance. Our research aims to address the challenges in human robot interaction including human attention allocation and optimization, human and robot joint decision making, learning based adaptation, and trust-based control. Principal investigator: Ningshi Yao
Reconstructing Firing Rate Dynamics in Brain Neural Networks
Every second, the human body sends 11 million bits of information to the brain for processing. Given the brain’s tremendous capability for information filtering and decision making, understanding its mechanism has become one of the most promising topics in order to boost the next generation of artificial intelligence. To study brain activities from a quantitative perspective, the firing rate of neurons (e.g., number of spikes per second) has become a widely adopted measure due to its trial-to-trial reproducibility in experiments. Motivated by this, our research aims to reconstruct the firing rate dynamics in brain neural networks based on experimental data. The key approaches we employ include non-linear system identification and data-driven control. Principle Investigator: Xuan Wang
Resilient Multi-agent Coordination
Cyber-attack resilience is a fundamental demand for multi-agent systems, especially when the system is under dynamic environments with unstable communication channels and quickly-varying data-streams. Challenged by the un-known attacking sources, traditional information-security approaches focus on protecting the data itself, but are not directly applicable to autonomous multi-agent systems, which involve not only data gathering but also data processing, information sharing and coordination. Motivated by this, our research aims to develop new algorithms that allows multi-agent system to achieve resilient coordination even in the presence of sophisticated cyber-attacks. Principle Investigator: Xuan Wang
Network Science and Distributed Control Theory
Rapid development of technology is quickly leading us to an increasingly networked and wireless world. With massive wireless networks on the horizon, the efficient coordination of such large networks becomes an important consideration that existing control theory is not equipped to handle alone. Given these new-age considerations, new fundamental tools and methods combining ideas from network science and decision/control theory are required to enable the reliable operation of large-scale networks of the future. General applications include networked Cyber-Physical Systems (CPS), the Internet of Things (IoT), multi-agent/swarm robotics, security and privacy of networks. Specific commercial applications include automated dispatch for ride-sharing vehicles or police cars, smart warehouses, smart grid, autonomous drone delivery systems, malware security in IoT networks, etc. Principal investigator: Cameron Nowzari.
SPARX – Swarming Platform for Autonomous Robots X
Utilizing swarms (very large numbers) of cheap/incapable robots is the ultimate goal here. Fish swim in schools of fish for protection, birds flock together for efficiency, and ants need the entire swarm to survive. Much like we see very sophisticated and seemingly intelligent behavior emerge from swarms of animals, we want to harness this power in the laboratory. To demonstrate the success of effectively utilizing a 'swarm' of robots, our robots are used in a twice-a-year aerial soccer game-type of competition hosted by the Office of Naval Research. Please see our project website for further details. Principle Investigator: Cameron Nowzari.