Research Topics

---

This page highlights some of the Personal Robotics research at Intel Labs Seattle:

RGBD (3D Vision)

We are exploring the use of new range cameras for machine perception.

• Multi-Sensor Object Recognition Using Web Data and Domain Adaptation
• Manipulator and Object Tracking for In-Hand Model Acquisition
• Semantic mapping: Building object-level maps of indoor environments

Novel Sensing for Robotics

How can new sensors simplify and advance robotics? We coined the term Pretouch to refer to in-hand senses that are shorter range than vision but longer range than touch.

• Electric Field Pretouch for Robotic Manipulation
• "Robot Feed Thyself": Autonomous robotic plugging in by sensing 60Hz emissions
• "WISP (Wireless Identification and Sensing Platform):" A battery-free sensing platform that is powered and read by standard UHF RFID readers

Other projects

• Human-Guided Manipulation: How can we transfer to robots the excellent manipulation capabilities of humans? This is a collaborative project with Prof. Yoky Matsuoka and her Neurobotics lab.
• Robot Security: What security risks and challenges do Personal Robots present? How do problems of robot security relate to ordinary computer security problems?
• Roadmapping robotics: Intel Personal Robotics researchers were selected to help draft the 15 year U.S. National Robotics Research roadmap. The cover of the roadmap document has in image from our research, an apple held by Marvin's e-field sensing hand. Our contribution to the workshop is in the publications page.
• Natural language interfaces to robots: How can robots learn to ground human language in their representations of their environments?