Research Areas

Distributed Control and Dynamical Systems

Distributed control and risk analysis of stochastic networked systems with applications to robot coordination and social networks. Most recent work includes visual perception and planning of robots.


Topological Methods in Robotics

Our research interests are centered around motion planning and control of autonomous, intelligent systems. More specifically, in applications of topological and geometric methods (algebraic/differential topology, differential/discrete geometry) to the design and analysis of algorithms in robot motion planning, coverage, sensor networks, distributed systems and control. In general, interested in diverse topics involving control theory, graph theory, topology, geometry and applied mathematics.


Robot Swarms

Robot swarms have the potential to self-recover from failures and self-adapt to changing environments. Cooperative work and modularity can also increase the capabilities of the robot swarm. We work on designing new robots and developing coordination algorithms, including modular aerial systems. Our main research areas are control for modular dynamical systems; self-reconfiguration and self-recovering algorithms, and distributed coordination for large multi-robot networks.


Robot Manipulation

For robots to be most useful, they have not only to sense the world around them, but also to manipulate objects. Manipulation is challenging because it often requires robots to make and break contact with objects and to control the contact forces that arise. For example, imagine a robot with a multi-fingered hand unscrewing a burned out lightbulb from a lamp. In our group, we design algorithms to produce control policies and motion plans for manipulation that learn from data while also leveraging traditional motion planning methods and physical models of contact. Our long-term goal is to develop methods that are adaptive to task uncertainties, extensible to new classes of tasks, and produce results that are explainable.


Smart Autonomous Systems

Faculty are involved in conducting research to build robust perception systems for autonomous systems and adding human-like features for home/office/warehouse robots. Autonomous systems typically are equipped with different types of sensors e.g. RGBD, LIDAR, ultrasonic sensors that are utilized in their perception systems. Designing efficient and robust perception systems enable such autonomous systems to navigate their complex environments more reliably. Similarly, artificial intelligence-related research enables us to add human-like features e.g. intelligent conversational agents that are useful for facilitating human-robot collaborations.