Intel Research's Personal Robotics project aims to develop useful robotic assistants for indoor, populated environments. To transition robots from structured, assembly line scenarios to more natural, unstructured environments requires advances in perception, reasoning under uncertainty, and human-robot interaction. We are currently pursuing research in all these areas - please see our publications page for more information.
The project is spread across our Pittsburgh and Seattle labs. The Pittsburgh team collaborates closely with the Quality of Life Technologies initiative at Carnegie Mellon University; the Seattle team collaborates with the University of Washington. Both teams have identical hardware. Currently, Pittsburgh focuses on mid-range perception, planning, and manipulation research, while Seattle focuses on short-range sensing and perception, and human-robot interaction. In the coming months, these efforts will be combined to provide an extremely competent mobile manipulation platform.
The current system in Pittsburgh combines a manipulator arm and hand with a mobile Segway for performing recognition, modeling, and manipulation of common indoor objects.
For more information, please see the various links to the left or feel free to contact either Dave or Sidd (email addresses: {dave.ferguson/siddhartha.srinivasa}@intel.com).
The Seattle system augments the Barrett hand with custom Electric Field Sensors in the fingertips to detect and model objects at close range without needing to touch them. For information on Electric Field Pretouch, contact joshua.r.smith@intel.com
