Online Adaptive Teleoperation via Incremental Intent Modeling
We present a task-independent framework that is amenable to long-duration teleoperation of a mobile robot in an unknown scenario, using incremental intent models. In particular, we focus on constructive teleoperation in performing an unknown task without knowledge of the environment, while allowing user to retain full control over the set of motions. We evaluate our method on a quadrotor flying in a lemniscate maneuver as a proof-of-concept experiment, and show reduced joystick entropy. We conclude with a discussion regarding our on-going investigation with this framework.
People
Xuning Yang
PhD, 2021. Now a Research Scientist in the Robotics Research Lab at NVIDIA Research
Nathan Michael
PI, 2014-2023. Now at Shield AI.
Robots
