Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Adaptive Human-Robot Collaborative Missions using Hybrid Task Planning

Gricel Vázquez, Alexandros Evangelidis, Sepeedeh Shahbeigi, Simos Gerasimou

TL;DR

This paper tackles robust task planning for cyber-physical-human systems under uncertainty by introducing a hybrid two-stage method that first generates a feasible plan via numerical planning and then augments it with probabilistic uncertainty modeling to synthesize Pareto-optimal, verified plans. It leverages a GA-driven search over task retry configurations and probabilistic model checking to balance mission success probability and cost, while supporting runtime adaptation through a MAPE-K loop. The vineyard case study demonstrates that the approach scales beyond full MDP verification, substantially reducing state explosion and enabling incremental replanning. The work offers practical impact for complex human-robot collaboration in dynamic environments by delivering verifiable plans that remain adaptable under changing requirements and uncertainties.

Abstract

Producing robust task plans in human-robot collaborative missions is a critical activity in order to increase the likelihood of these missions completing successfully. Despite the broad research body in the area, which considers different classes of constraints and uncertainties, its applicability is confined to relatively simple problems that can be comfortably addressed by the underpinning mathematically-based or heuristic-driven solver engines. In this paper, we introduce a hybrid approach that effectively solves the task planning problem by decomposing it into two intertwined parts, starting with the identification of a feasible plan and followed by its uncertainty augmentation and verification yielding a set of Pareto optimal plans. To enhance its robustness, adaptation tactics are devised for the evolving system requirements and agents' capabilities. We demonstrate our approach through an industrial case study involving workers and robots undertaking activities within a vineyard, showcasing the benefits of our hybrid approach both in the generation of feasible solutions and scalability compared to native planners.

Adaptive Human-Robot Collaborative Missions using Hybrid Task Planning

TL;DR

This paper tackles robust task planning for cyber-physical-human systems under uncertainty by introducing a hybrid two-stage method that first generates a feasible plan via numerical planning and then augments it with probabilistic uncertainty modeling to synthesize Pareto-optimal, verified plans. It leverages a GA-driven search over task retry configurations and probabilistic model checking to balance mission success probability and cost, while supporting runtime adaptation through a MAPE-K loop. The vineyard case study demonstrates that the approach scales beyond full MDP verification, substantially reducing state explosion and enabling incremental replanning. The work offers practical impact for complex human-robot collaboration in dynamic environments by delivering verifiable plans that remain adaptable under changing requirements and uncertainties.

Abstract

Producing robust task plans in human-robot collaborative missions is a critical activity in order to increase the likelihood of these missions completing successfully. Despite the broad research body in the area, which considers different classes of constraints and uncertainties, its applicability is confined to relatively simple problems that can be comfortably addressed by the underpinning mathematically-based or heuristic-driven solver engines. In this paper, we introduce a hybrid approach that effectively solves the task planning problem by decomposing it into two intertwined parts, starting with the identification of a feasible plan and followed by its uncertainty augmentation and verification yielding a set of Pareto optimal plans. To enhance its robustness, adaptation tactics are devised for the evolving system requirements and agents' capabilities. We demonstrate our approach through an industrial case study involving workers and robots undertaking activities within a vineyard, showcasing the benefits of our hybrid approach both in the generation of feasible solutions and scalability compared to native planners.

Paper Structure

This paper contains 15 sections, 4 equations, 6 figures, 4 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Vineyard Inspection CPHS Mission
  3. Background
  4. Numeric planning
  5. Probabilistic Model Checking
  6. Problem Formulation
  7. Adaptive and Hybrid Task Planning
  8. Overview
  9. Approach
  10. Evaluation
  11. Research questions
  12. Experimental setup
  13. Results and Discussion
  14. Related work
  15. Conclusions and Future Work

Figures (6)

  • Figure 1: Vineyard layout and abstract representation. Locations $l_1$-$l_9$ represent a section of a vineyard's row, where tasks $t1$-$t3$ have to be completed. Task $t1$ can be done by human workers, $t2$ by robots and $t3$ by either.
  • Figure 2: Overview of our approach showing the required steps and generated artefacts, and adaptation relevant parts.
  • Figure 3: Input file (a), generated PDDL files (b,c), parametric DTMC with probabilistic properties (d) and synthesis Pareto-optimal set of solutions (plans). The highlighted sections of the PDDL files use a colour palette based on the corresponding parts of the input file. The yellow parts refer to the task planning with uncertainty quantification part (see Figure \ref{['fig:overview']}).
  • Figure 4: Pareto front from optimal baseline and our hybrid solution for problem instances. From left to right:$\mathcal{M}_1, \mathcal{M}_2, \mathcal{M}_3, \mathcal{M}_4$.
  • Figure 5: Sample scenario showing adaptation cases A1-A3.
  • ...and 1 more figures

Theorems & Definitions (3)

  • Definition 1: Problem Specification
  • Definition 2: Problem Solution
  • Definition 3: Pareto Front Solutions