SynthCharge: An Electric Vehicle Routing Instance Generator with Feasibility Screening to Enable Learning-Based Optimization and Benchmarking

2026-03-03

AI Summary

This work addresses the limitations of existing Electric Vehicle Routing Problem with Time Windows (EVRPTW) benchmark datasets, which are predominantly static and lack rigorous feasibility validation, thereby hindering reproducible evaluation of learning-based optimization methods. To overcome this, we propose SynthCharge—a parameterized instance generator that uniquely integrates instance geometry, adaptive battery capacity scaling, and range-aware charging station placement. By incorporating an efficient feasibility filtering mechanism, SynthCharge ensures the validity of generated instances while supporting diverse and structurally coherent problem configurations. The framework enables the generation of EVRPTW instances spanning 5 to 100 customer nodes, offering a dynamic, scalable, and reproducible benchmarking infrastructure for neural routing and data-driven algorithmic approaches.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Search and Optimization: Sampling/Simulation-based Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments

Abstract

The electric vehicle routing problem with time windows (EVRPTW) extends the classical VRPTW by introducing battery capacity constraints and charging station decisions. Existing benchmark datasets are often static and lack verifiable feasibility, which restricts reproducible evaluation of learning-based routing models. We introduce SynthCharge, a parametric generator that produces diverse, feasibility-screened EVRPTW instances across varying spatiotemporal configurations and scalable customer counts. While SynthCharge can currently generate large-scale instances of up to 500 customers, we focus our experiments on sizes ranging from 5 to 100 customers. Unlike static benchmark suites, SynthCharge integrates instance geometry with adaptive energy capacity scaling and range-aware charging station placement. To guarantee structural validity, the generator systematically filters out unsolvable instances through a fast feasibility screening process. Ultimately, SynthCharge provides the dynamic benchmarking infrastructure needed to systematically evaluate the robustness of emerging neural routing and data-driven approaches.

Problem

• electric vehicle routing problem
• time windows
• benchmark dataset
• feasibility
• learning-based optimization

Innovation

• EVRPTW
• instance generation
• feasibility screening
• learning-based optimization
• benchmarking