Deep Learning--Accelerated Multi-Start Large Neighborhood Search for Real-time Freight Bundling

2025-12-11

AI Summary

This work addresses real-time freight order aggregation in Online Freight Exchange (OFEX) systems, formulated as the multi-commodity one-to-one selective Pickup-and-Delivery Traveling Salesman Problem (m1-PDSTSP), aiming to maximize revenue under capacity, precedence, and path-length constraints. We propose the first Transformer-based constructive heuristic to generate high-quality initial solutions for Multi-Start Large Neighborhood Search (MSLNS). Crucially, we provide the first empirical validation that learned constructive heuristics consistently enhance the performance of improvement heuristics—and generalize effectively across diverse selective TSP and pickup-and-delivery variants. Integrated within a rolling time-window framework, our approach achieves sub-second solving latency and less than 2% optimality gap on benchmark instances, significantly outperforming state-of-the-art neural combinatorial optimization methods and conventional metaheuristics.

Technology Category

• Search and Optimization: Combinatorial Optimization
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Online Freight Exchange Systems (OFEX) play a crucial role in modern freight logistics by facilitating real-time matching between shippers and carrier. However, efficient combinatorial bundling of transporation jobs remains a bottleneck. We model the OFEX combinatorial bundling problem as a multi-commodity one-to-one pickup-and-delivery selective traveling salesperson problem (m1-PDSTSP), which optimizes revenue-driven freight bundling under capacity, precedence, and route-length constraints. The key challenge is to couple combinatorial bundle selection with pickup-and-delivery routing under sub-second latency. We propose a learning--accelerated hybrid search pipeline that pairs a Transformer Neural Network-based constructive policy with an innovative Multi-Start Large Neighborhood Search (MSLNS) metaheuristic within a rolling-horizon scheme in which the platform repeatedly freezes the current marketplace into a static snapshot and solves it under a short time budget. This pairing leverages the low-latency, high-quality inference of the learning-based constructor alongside the robustness of improvement search; the multi-start design and plausible seeds help LNS to explore the solution space more efficiently. Across benchmarks, our method outperforms state-of-the-art neural combinatorial optimization and metaheuristic baselines in solution quality with comparable time, achieving an optimality gap of less than 2% in total revenue relative to the best available exact baseline method. To our knowledge, this is the first work to establish that a Deep Neural Network-based constructor can reliably provide high-quality seeds for (multi-start) improvement heuristics, with applicability beyond the extit{m1-PDSTSP} to a broad class of selective traveling salesperson problems and pickup and delivery problems.

Problem

• Optimizes revenue-driven freight bundling under constraints
• Couples bundle selection with routing under sub-second latency
• Provides high-quality seeds for improvement heuristics in combinatorial problems

Innovation

• Transformer Neural Network-based constructive policy for initial solutions
• Multi-Start Large Neighborhood Search metaheuristic for iterative improvement
• Rolling-horizon scheme to handle real-time freight bundling efficiently