Assistive Robot Teleoperation Using Behavior Trees

TL;DR

This paper introduces a method for assistive robot teleoperation that uses Behavior Trees with shared control nodes to improve user control, obstacle avoidance, and activity representation in noisy or delayed environments.

Contribution

It presents a novel approach combining Behavior Trees with shared control nodes for flexible, user-friendly activity representation and assistance in robot teleoperation.

Findings

Behavior Trees enable easy activity definition by end users.

Shared Control Action Nodes improve task execution and obstacle avoidance.

The approach handles noisy and delayed input signals effectively.

Abstract

Robotic assistance in robot arm teleoperation tasks has recently gained a lot of traction in industrial and domestic environment. A wide variety of input devices is used in such setups. Due to the noise in the input signals (e.g., Brain Computer Interface (BCI)) or delays due to environmental conditions (e.g., space robot teleoperation), users need assistive autonomy that keeps them in control while following predefined trajectories and avoids obstacles. This assistance calls for activity representations that are easy to define by the operator and able to take the dynamic world state into consideration. This paper represents Activities of Daily Living using Behavior Trees (BTs) whose inherent readability and modularity enables an end user to define new activities using a simple interface. To achieve this, we augment BTs with Shared Control Action Nodes, which guide the user's input on a trajectory facilitating and ensuring task execution.