Informing Real-Time Corrections in Corrective Shared Autonomy Through Expert Demonstrations

@article{Hagenow2021InformingRC,
  title={Informing Real-Time Corrections in Corrective Shared Autonomy Through Expert Demonstrations},
  author={Michael Hagenow and Emmanuel Senft and Robert G. Radwin and Michael Gleicher and Bilge Mutlu and Michael R. Zinn},
  journal={IEEE Robotics and Automation Letters},
  year={2021},
  volume={6},
  pages={6442-6449}
}
Corrective Shared Autonomy is a method where human corrections are layered on top of an otherwise autonomous robot behavior. Specifically, a Corrective Shared Autonomy system leverages an external controller to allow corrections across a range of task variables (e.g., spinning speed of a tool, applied force, path) to address the specific needs of a task. However, this inherent flexibility makes the choice of what corrections to allow at any given instant difficult to determine. This choice of… 

Figures and Tables from this paper

Affordance Template Registration via Human-in-the-loop Corrections
TLDR
This paper proposes a registration method which combines autonomy and user corrections, and presents an overview of existing methods, a description of the method, preliminary results, and planned future work.

References

SHOWING 1-10 OF 20 REFERENCES
Corrective Shared Autonomy for Addressing Task Variability
TLDR
This letter presents corrective shared autonomy, where users provide corrections to key robot state variables on top of an otherwise autonomous task model, and demonstrates its viability and benefits such as low user effort and physical demand via a system-level user study on three tasks involving variability situated in aircraft manufacturing.
A learning-based shared control architecture for interactive task execution
TLDR
The goal of this paper is to present a shared control framework that uses learned expert distributions to gain more autonomy, resulting in a master-slave system with increasing autonomy that requires less user input with an increasing number of task executions.
Learning the Correct Robot Trajectory in Real-Time from Physical Human Interactions
TLDR
This work explores a one-shot approach that enables robots to harness physical human interventions to update their trajectory and goal during autonomous tasks without relying on multiple iterations, offline demonstrations, or replanning.
Perception-Aware Human-Assisted Navigation of Mobile Robots on Persistent Trajectories
We propose a novel shared control and active perception framework combining the skills of a human operator in accomplishing complex tasks with the capabilities of a mobile robot in autonomously
Semi-autonomous trajectory generation for mobile robots with integral haptic shared control
TLDR
A new framework for semi-autonomous path planning for mobile robots that extends the classical paradigm of bilateral shared control is presented and is validated with extensive experiments using a quadrotor UAV and a human in the loop with two haptic interfaces.
Learning Control
  • E. Grant
  • Computer Science
    Encyclopedia of Machine Learning
  • 2010
TLDR
The probabilistic forward dynamics models can be employed to control complex musculoskeletal robots on an antagonistic pair of pneumatic artificial muscles using only one-step-ahead predictions of the forward model and incorporating model uncertainty.
A survey of robot learning from demonstration
Learning from Interventions: Human-robot interaction as both explicit and implicit feedback
TLDR
This work argues that learning interactively from expert interventions enjoys the best of both worlds, and formalizes this as a constraint on the learner’s value function, which it can efficiently learn using no regret, online learning techniques.
Using learning from demonstration to generate real-time guidance for haptic shared control
This paper introduces a new Learning from Demonstration (LfD)-based method that makes usage of robot effector forces and torques recorded during expert demonstrations, to generate force-based haptic
Robot trajectory modification using human-robot force interaction
  • Hung-Shen LiuK. Song
  • Computer Science
    2017 56th Annual Conference of the Society of Instrument and Control Engineers of Japan (SICE)
  • 2017
TLDR
This paper presents a method to teach a humanoid robot arm a new trajectory by using human-robot shared control and proposed method to modify part of the trajectory based on shared control to resolve this problem.
...
...