Carnegie Mellon Robotics Institute
I am interested in making robots act purposefully and successfully in a world in which most everything is uncertain. Sensors are noisy, actions are imprecise, and objects are often in the wrong location. Despite such obstacles to purposeful action, there are many tasks that can be accomplished successfully. Humans, animals, and some machines are proof. Providing robots with the ability to operate autonomously and purposefully requires an understanding of how different tasks may be accomplished by different repertoires of actions. Grasping, hitting, and dropping are some actions that are useful in a robot's repertoire. More exotic actions include shaking, twirling, and other actions that randomize an object's state. Recently I constructed a "two-palm robot". The robot consisted of two manipulator arms cooperating to manipulate objects without the need for full kinematic constraint. The arms "programmed themselves", that is, they invoked an automatic planner to find sequences of motions for reorienting objects in their palms. The planner built a geometric graph based on a critical event analysis of the underlying mechanics.
My work is motivated by several desires. First, I would like to program robots more easily than is currently possible. Second, I would like to understand the scope and limitations of autonomous systems, whether biological or artificial. Third, I would like to reduce the complexity of design and planning by codifying the design parameters required to achieve a given level of automation. An underlying goal of my research is to understand the relationship between sensing, action, and prediction. In the past, I have explored various extreme points in this space. With Matt Mason I explored sensorless strategies, for my thesis work I looked at randomized strategies, and most recently I investigated fast-action minimal-sensing strategies. My research draws on tools from geometry, mechanics, planning, and stochastic processes.
I am interested in sensing strategies that acquire object shape and configuration concurrently during manipulation, a research direction pioneered by my former students Yan-Bin Jia and Mark Moll. Currently, Matt Mason, Sidd Srinivasa, and I are working on a related project to develop a theory of task-level kinesthetic perception.
I am also interested in protein homology, in particular determining structural homology from sparse NMR data. I am exploring techniques from knot theory to model protein structures.
|Research Interest Keywords|
|manipulation mechanics, mobile manipulation, motion planning, protein knot theory, protein structure, shape sensing|
|The Robotics Institute is part of the School of Computer Science, Carnegie Mellon University.|
Contact Us | Update Instructions