Curating User-Defined Interface Maps for Robot Teleoperation

Abstract

The interfaces used to operate assistive robots typically employ fixed, predefined maps to associate interfacelevel commands to robot control commands. User-defined control maps instead consider an individual's preferences and capabilities, moving away from a one-size-fits-all mapping paradigm. This work presents novel methods for (1) eliciting user-defined control maps, (2) identifying and addressing issues in control signal data that arise from such a user-centered design, and (3) methodologies to filter erroneous signals and construct synthetic data in an effort to address these issues. We experimentally evaluate our proposed methods by conducting a user study that elicits user-defined, interface-level commands for controlling a powered wheelchair and a robotic arm through four control interfaces. Our results highlight the differing suitability of user-defined, data-driven control maps for the various interface-platform pairings, and provide an analysis of the errors generated during dataset definition and the impact of our postprocessing methods.