Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Online Resynthesis of High-Level Collaborative Tasks for Robots with Changing Capabilities

Amy Fang, Tenny Yin, Hadas Kress-Gazit

TL;DR

The paper tackles online resynthesis for a heterogeneous robot team executing a global task specified in $LTL^{ψ}$ when individual capabilities change during execution. It extends the $LTL^{ψ}$ grammar with binding constraints $c_{distinct}$ and $c_{min}$ and proposes a hierarchical resynthesis framework that minimizes reallocation by performing local updates to per-robot product automata. The approach maintains task satisfaction by updating models and coordinating binding reallocation, and it can fall back to a full team replan if needed. A simulated warehouse demonstration demonstrates robustness to failures/additions and quantifies computation times, highlighting practical feasibility.

Abstract

Given a collaborative high-level task and a team of heterogeneous robots and behaviors to satisfy it, this work focuses on the challenge of automatically, at runtime, adjusting the individual robot behaviors such that the task is still satisfied, when robots encounter changes to their abilities--either failures or additional actions they can perform. We consider tasks encoded in LTL^ψand minimize global teaming reassignments (and as a result, local resynthesis) when robots' capabilities change. We also increase the expressivity of LTL^ψby including additional types of constraints on the overall teaming assignment that the user can specify, such as the minimum number of robots required for each assignment. We demonstrate the framework in a simulated warehouse scenario.

Online Resynthesis of High-Level Collaborative Tasks for Robots with Changing Capabilities

TL;DR

The paper tackles online resynthesis for a heterogeneous robot team executing a global task specified in when individual capabilities change during execution. It extends the grammar with binding constraints and and proposes a hierarchical resynthesis framework that minimizes reallocation by performing local updates to per-robot product automata. The approach maintains task satisfaction by updating models and coordinating binding reallocation, and it can fall back to a full team replan if needed. A simulated warehouse demonstration demonstrates robustness to failures/additions and quantifies computation times, highlighting practical feasibility.

Abstract

Given a collaborative high-level task and a team of heterogeneous robots and behaviors to satisfy it, this work focuses on the challenge of automatically, at runtime, adjusting the individual robot behaviors such that the task is still satisfied, when robots encounter changes to their abilities--either failures or additional actions they can perform. We consider tasks encoded in LTL^ψand minimize global teaming reassignments (and as a result, local resynthesis) when robots' capabilities change. We also increase the expressivity of LTL^ψby including additional types of constraints on the overall teaming assignment that the user can specify, such as the minimum number of robots required for each assignment. We demonstrate the framework in a simulated warehouse scenario.
Paper Structure (21 sections, 5 equations, 4 figures, 1 algorithm)

This paper contains 21 sections, 5 equations, 4 figures, 1 algorithm.

Table of Contents

  1. Introduction
  2. Task Grammar: LTL$^\psi$
  3. Prior work - Robot Model and Büchi Automaton
  4. Robot Model
  5. Büchi Automaton for an LTL$^\psi$ Formula
  6. Behavior Synthesis
  7. Problem Setup
  8. Modifications
  9. Problem Statement
  10. Example
  11. Approach: Resynthesis Framework
  12. Evaluating Modified Robot's Behavior
  13. Updating the Product Automaton
  14. Constructing $A_j^{mod}$ with $\Delta_j^{rem}$
  15. Constructing $A_j^{mod}$ with $\Delta_j^{add}$
  16. ...and 6 more sections

Figures (4)

  • Figure 1: Modifications to $\lambda_{mot}$ in which (a) the robot can no longer move between rooms G and H ($\Delta_{mot}^{rem}$ is the set of red transitions), and (b) the robot can now move between rooms B and G ($\Delta_{mot}^{add}$ is the set of green transitions).
  • Figure 2: Environment and robot setup
  • Figure 3: Buchi Automaton $\mathcal{B}$ for the example in Sec. \ref{['sec:ex']}. The highlighted transitions is $\beta$, the trace that the team of robots are collectively traversing to satisfy the task.
  • Figure 4: Overview of resynthesis process

Theorems & Definitions (2)

  • Definition 1: Capability Function
  • Definition 2: Binding Assignment Function