My interests span a range of issues that are relevant to building autonomous agents including perception, motion planning, task planning, manipulation, learning, and human-robot interaction. I believe that the constraints imposed by the physical environment and by the demands of human users are important for grounding robotics research. I like to develop general-purpose solutions that have a strong mathematical basis and perform comprehensive evaluation on hardware platforms. I enjoy field-work with real-world robots and have experience working with autonomous under-water vehicles, mobile robots, and manipulator arms.

For my dissertation research, I recognized that for an assistive robot (such as an intelligent wheelchair) to be acceptable to human users, its motion should be safe and comfortable. Further, different users should be able to customize the motion according to their comfort. My work formalizes the notion of motion comfort as a discomfort measure that can be minimized to compute comfortable trajectories, and identifies several properties that a trajectory must have for the motion to be comfortable. Further, I developed motion planning framework for planning safe, comfortable, and customizable motion of wheeled mobile robots moving on a plane. This framework computes a trajectory, i.e., a function of robot configuration with respect to time. A trajectory encodes (i) an obstacle-free free geometric path, and (ii) a kinodynamically feasible motion on the path. All possible motion tasks with different boundary conditions on speed and acceleration are handled (e.g. starting from rest, coming to rest, continuing to move). This framework includes star-shaped obstacles, kinematic and dynamic constraints resulting from physical limitations and comfort requirements, and allows the user to customize the motion for comfort.

To the best of my knowledge, this is the first work that addresses the problem of planning comfortable and customizable motion for assistive mobile robots. Our framework removes the limitations of existing motion planning methods, none of which can plan trajectories that have all the properties necessary for comfort. To the best of our knowledge, this is the first comprehensive formulation of kinodynamic motion planning for wheeled mobile robots that includes all of the following -- a careful analysis of boundary conditions and continuity requirements on trajectory, dynamic constraints, obstacle avoidance constraints, and a robust numerical method that computes solution trajectories in a few seconds.

This motion-planning framework is not limited to assistive robots. It applies equally well to motion planning of other wheeled robots including robotic cars.

I was part of the research, software, and field team for ENDURANCE, a NASA-funded project. ENDURANCE is a hovering AUV that was developed by Stone Aerospace as a platform for testing technologies for discovery of life form on watery moons such as Europa. ENDURANCE performed extensive scientific exploration of West Lake Bonney in the Antarctic Dry Valleys in austral summer of 2008 and 2009. I, along with colleagues, developed a novel vision-based homing algorithm for recovering the vehicle through a melt hole in the lake ice.

Previously, I worked on developing algorithms and implementing software for supervisory control and task planning of semi-automated tele-operated robotic surgery cell.