VRPAgent: LLM-Driven Discovery of Heuristic Operators for Vehicle Routing Problems

2025-10-08

AI Summary

Vehicle Routing Problems (VRPs) have long relied on expert-crafted heuristics, with automated heuristic design remaining a longstanding challenge. This paper introduces the first framework leveraging large language models (LLMs) to generate evolvable, problem-specific heuristic operators. These LLM-generated operators are embedded within a generic metaheuristic architecture and jointly optimized end-to-end via genetic search. The framework enforces functional correctness and search efficacy of the synthesized operators through integrated validation and fitness-driven evolution. Empirical evaluation across diverse VRP benchmarks—including capacitated VRP, VRP with time windows, and orienteering problems—demonstrates consistent and significant improvements over both handcrafted heuristics and state-of-the-art learning-based baselines. Notably, the method operates efficiently on a single CPU core, eliminating reliance on specialized hardware. This work establishes a novel paradigm for automated heuristic discovery in combinatorial optimization, bridging foundation models and evolutionary computation.

Technology Category

• Search and Optimization: Heuristic Search
• Planning, Routing, and Scheduling: Planning with Language Models
• Intelligent Robots: Learning & Optimization for ROB

Application Category

• Search and Retrieval-Augmented AI: Large language models for search
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Designing high-performing heuristics for vehicle routing problems (VRPs) is a complex task that requires both intuition and deep domain knowledge. Large language model (LLM)-based code generation has recently shown promise across many domains, but it still falls short of producing heuristics that rival those crafted by human experts. In this paper, we propose VRPAgent, a framework that integrates LLM-generated components into a metaheuristic and refines them through a novel genetic search. By using the LLM to generate problem-specific operators, embedded within a generic metaheuristic framework, VRPAgent keeps tasks manageable, guarantees correctness, and still enables the discovery of novel and powerful strategies. Across multiple problems, including the capacitated VRP, the VRP with time windows, and the prize-collecting VRP, our method discovers heuristic operators that outperform handcrafted methods and recent learning-based approaches while requiring only a single CPU core. To our knowledge, VRPAgent is the first LLM-based paradigm to advance the state-of-the-art in VRPs, highlighting a promising future for automated heuristics discovery.

Problem

• Automating heuristic design for vehicle routing problems
• Generating problem-specific operators using large language models
• Improving performance over handcrafted and learning-based methods

Innovation

• LLM generates problem-specific heuristic operators
• Integrates operators into metaheuristic with genetic search
• Discovers novel strategies outperforming handcrafted methods