Research

Control and Learning for Autonomous Robots

Recent years have seen a massive increase in applications of autonomous robots. ICON researchers are pushing the frontiers of robotics by addressing emerging challenges arising from unstructured environment planning and manipulation, whole-body locomotion on dynamic terrain, coordination of large number of autonomous robots and so on. Besides providing fundamental solutions from integration of control and learning, ICON researchers also emphasize applications of individual robotic arms, humanoid robots, soft robots as well as robotic swarms in manufacturing, medical services, precision agriculture and space logistics.

Resilient and Trusted Autonomy

With society's increasing dependence on autonomous systems, it becomes critical to mitigate the impacts of failures and malicious attacks on those systems. Meanwhile these autonomous systems must be trustworthy. Resilient and trusted autonomy must be designed into such systems, allowing them to adapt to sophisticated attacks and failures while being trusted to meet their objectives. ICON researchers are developing innovative approaches that blend foundational theory with recent advances in AI and learning to enable autonomous systems to achieve resiliency, safety, and trust.

Human-Autonomy Teaming

The future of autonomy demands cooperative and seamless interaction between it and humans across a range of sectors, from healthcare to the military theater. ICON faculty, working closely with researchers in human factors, cognitive psychology, and behavioral economics, are taking an interdisciplinary approach toward developing and validating the necessary theory and algorithms that are going to realize the promise of truly synergistic cyber-physical and human systems.

Multi-Agent and Networked Systems

ICON researchers are tackling the challenging problems of control, optimization, and learning for interconnected and multi-agent systems, with a wide variety of applications including epidemic monitoring and control, robotic swarms, and cyber-physical systems. Algorithms are being developed to allow such networked systems to work as a cohesive whole, accounting for complex dynamics and time-varying interactions between the various entities in the network.

Intelligence for Infrastructure Systems

Modern infrastructure systems are tightly coupled with each other and with the social systems that they serve. Interdisciplinary teams of ICON researchers are laying the foundations for designing, optimizing, and controlling these complex systems at a variety of scales, ranging from the individual subsystems to the overall society-scale system. Key application areas include the smart grid, intelligent transportation systems, precision agriculture, advanced manufacturing, and disaster-resilient communities.

Validation and Verification of Autonomy

Transitioning research from theory to practice requires techniques to verify and validate the performance of autonomous and connected vehicles (drones, UAS, ground robots, cars, etc.) before deployment. A variety of cutting-edge labs and facilities at ICON enable such validation and verification of systems, ranging from single autonomous vehicles to large-scale connected systems.