World embedded interfaces for human-robot interaction

Abstract

Human interaction with large numbers of robots or distributed sensors presents a number of difficult challenges including supervisory management, monitoring of individual and collective state, and apprehending situation awareness. A rich source of information about the environment can be provided even with robots that have no explicit representations or maps of their locale. To do this, we transform a robot swarm into a distributed interface embedded within the environment. Visually, each robot acts like a pixel within a much larger visual display space so that any robot need only communicate a small amount of information from its current location. Our approach uses augmented reality techniques for communicating information to humans from large numbers of small-scale robots to enable situation awareness, monitoring, and control for surveillance, reconnaissance, hazard detection, and path finding.