Research
Home Research Publications Courses People Pictures Links Join Us
Nonprehensile Mobile Manipulation  We are exploring problems of manipulation, mobility, and cooperation using mobile robots equipped with simple nonprehensile (nongrasping) manipulators. A single robot with such a manipulator has limited manipulation capabilities, but in cooperation with other robots, it becomes a capable robotic system.
Tapping Micropositioning Cell  One strategy for fine positioning is to tap an object, giving it some initial velocities, and letting it slide until it comes to rest due to friction. By repeating this process, objects can be positioned very precisely. We are building a robotic system to take advantage of this mode of manipulation.
Robot Coverage  Many applications require a sensor or actuator to be passed over a region. Examples include mowing a lawn, spray painting, and search & rescue. How should a single robot accomplish this task efficiently? How should multiple robots coordinate?
Multiple Robot Coordination  We are investigating the coordination and motion planning of large numbers of multiple robots. This is computationally hard when coordinating tens or hundreds or robots. Applications include the collision-free coordination of manipulators in automotive workcells, AGVs in factories and at loading docks, and potentially, interactive graphics characters or digital choreography.
Manipulation of Articulated Objects  Manipulating nonrigid objects, such as pizza cartons, is a challenging problem. We are developing techniques for the robotic manipulation, motion planning, and design of articulated objects with many degrees of freedom. The challenge is to enable robots with far fewer degrees of freedom to manipulate such objects. Potential applications include carton folding for automated packaging and bulk fabrication of 3D microelectromechanical systems (MEMS) devices from 2D hinged elements.
Parts Feeding  Parts feeding is the process of orienting and transferring parts for an assembly from their initial random configurations. We use knowledge of the mechanics and geometry of tasks to automatically generate parts feeding strategies.
 
Send mail to roboweb@cs.rpi.edu with questions or comments about this web site.
Copyright © 2000 Rensselaer Polytechnic Institute
Last modified: March 16, 2001