|Advanced Teleoperation Interfaces|
Easy-to-use tools for efficient, robust vehicle teleoperation in unknown and unstructured environments will enable joint human-robot inspection, maintenance, assembly and other critical in-space and surface activities.|
The Computational Sciences Division at NASA Ames Research Center in collaboration with the Robotics Institute (Carnegie Mellon University) and the Virtual Reality and Active Interfaces Group (Swiss Federal Institute of Technology/Lausanne) provide advanced teleoperation interfaces capabilities for performing critical joint human-robot tasks
Image left: Humans do not have to dedicate continuous attention to the robot. Image courtesy: NASA.
Since 2000, we have been developing a new human-robot interaction model, collaborative control, which is based on human-robot dialogue. Instead of a supervisor dictating to a subordinate, the human and the robot converse to exchange information, to ask questions, and to resolve differences. With this approach, human-robot interaction is more natural, balanced, and direct than conventional approaches. With collaborative control, the human functions as a resource for the robot. In particular, the robot is allowed to ask the human questions as it works, to obtain assistance with perception and cognition. For example, the robot might ask "Based on this image, is it safe to drive forward?" This allows the human to compensate for limited autonomy, but does not force him to dedicate continuous attention to the robot.
Visual Gestures and Haptics
Image right: Hand motions will direct the robot. Image courtesy: NASA.
GestureDriver (2000) is a remote driving interface based on visual gesturing. Hand motions are tracked with a color and stereo vision system and classified into gestures using a simple geometric model. The gestures are then mapped into motion commands, which are transmitted to the robot for execution. HapticDriver (2000) uses force-feedback to improve precision driving, such as required for maneuvering in cluttered spaces or when approaching a docking target. Information from a robot's range sensors is transformed to spatial forces using a linear model and then displayed with a large-workspace haptic hand controller (the Delta Haptic Device). In essence, this enables the user to feel the remote environment.
Non-Invasive Human-Robot Interfaces
Non-invasive interfaces enable humans to more naturally communicate with robots in proximity and long-distance interaction. Research success in this regard includes both voice-based interaction and dialogue management as well as indirect muscular sensing. Electromyographic techniques do not require physical contact between sensors, which can be sewn into EVA suits of Astronaut operators for instance.