M^3RS: Multi-robot, Multi-objective, and Multi-mode Routing and Scheduling

2024-03-24

arXiv.org

AI Summary

Existing multi-robot task allocation frameworks largely neglect Quality of Service (QoS) as a critical decision variable, hindering their applicability to time-sensitive, quality-critical tasks such as disinfection and cleaning. Method: This paper introduces the first QoS-aware framework where QoS is modeled as an adjustable decision variable, enabling multiple execution modes per task—each characterized by distinct resource consumption, time cost, and quality level—and jointly optimizing robot paths, task sequences, and mode selections. We propose a multi-robot, multi-objective, multi-mode scheduling framework that treats QoS as a first-class optimization objective and develop a clustering-enhanced column generation (CCG) algorithm to ensure both solution accuracy and computational efficiency. Results: Experiments in disinfection services demonstrate simultaneous optimization of sterilization quality and task completion rate; CCG achieves 2.5× speedup over baseline MIP solvers and exhibits strong robustness and practicality, validated via both simulation and real-robot deployment.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Intelligent Robots: Multi-Robot Systems
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

The quality of task execution can significantly impact a multi-robot mission. While higher quality is desirable, it may not always be feasible due to mission constraints. Existing multi-robot task allocation literature generally overlooks quality of service as a decision variable. Addressing this gap, we introduce the multi-robot, multi-objective, and multi-mode routing and scheduling (M^3RS) problem, designed for time-bound, multi-robot, multi-objective missions. In M^3RS, each task offers multiple execution modes, each with different resource requirements, execution time, and quality. M^3RS optimizes task sequences and execution modes for each agent. The need for M^3RS comes from multi-robot applications in which a trade-off between multiple criteria can be achieved by varying the task level quality of service through task execution modes. Such ability is particularly useful for service robot applications. We use M^3RS for the application of multi-robot disinfection in healthcare environments and other public locations. The objectives considered for disinfection application are disinfection quality and number of tasks completed. A mixed-integer linear programming (MIP) model is proposed for M^3RS. Further, a clustering-based column generation (CCG) algorithm is proposed to handle larger problem instances. Through synthetic, simulated, and hardware case studies, we demonstrate the advantages of M^3RS, showing it provides flexibility and strong performance across multiple metrics. Our CCG algorithm generates solutions 2.5x faster than a baseline MIP optimizer, maintaining competitive performance. The videos for the experiments are available on the project website: https://sites.google.com/view/g-robot/m3rs/

Problem

• Optimizing multi-robot task allocation with quality trade-offs
• Scheduling tasks with multiple execution modes and objectives
• Improving disinfection quality and efficiency in constrained missions

Innovation

• Multi-mode task execution for quality trade-offs
• Mixed-integer linear programming optimizes scheduling
• Clustering-based column generation reduces computation time