Toward Multi User Knowledge Driven Teleoperation of a Robotic Team with Scalable Autonomy

Abstract

This paper proposes a knowledge-driven teleoperation framework that enables multiple operators to command a team of robots to execute complex tasks in an efficient and intuitive manner. The framework leverages a shared knowledge base that captures domain-specific information and procedural knowledge about the task at hand. This knowledge base is used by a hybrid planner to generate context-specifically relevant commands for supervised autonomy robot command as well as direct teleoperation modes. By filtering the available commands, the operators are guided in their decision-making towards efficient task completion. This paper further extends our knowledge driven approach to address the switching between multiple operators and robotic assets, with the aim to be able scale up human-robot team for space exploration. Overall, this work represents a step towards more intelligent and collaborative teleoperation systems. The described system will be used in the Surface Avatar ISS-to-ground experiments slated for July 2023.