Design and Evaluation of a Multi-Robot Control Interface

Abstract

Designing a human-machine interface for a semi-autonomous, mobile, multi-robot system is a challenging task. The requirements include operating in a real-time environment, facilitating asynchronous command execution, and supporting the operator in dividing his or her monitoring and control resources among multiple robots. This paper presents the results of two simulation-based, multi-robot experiments that were conducted to guide and support the development of a multi-robot control interface. The experiments were designed around a simple mission-independent task of navigating robots towards goal points. While working with an identical setup, the two experiments differed in the level of autonomy provided by the robots. In the first experiment, the robots' navigational capabilities were realized by a simple line-of-sight algorithm requiring the operator to explicitly specify a path. In the second experiment, a more elaborate path generation algorithm was used, reducing the operator's activity to the selection of a goal point. In separate trials, operators had to manage 2, 4, and 8 robots in the two different environments. Robot autonomy represents a requirement for a multi-robot system managed by a single operator. Understanding the impact of varying levels of robot autonomy on operator performance is important for the interface design process. (9 figures, 15 refs.)