Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

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

Mertcan Daysalilar, Fuat Uyguroglu, Gabriel Nicolosi, Adam Meyers

TL;DR

SynthCharge is introduced, a parametric generator that produces diverse, feasibility-screened EVRPTW instances across varying spatiotemporal configurations and scalable customer counts, and provides the dynamic benchmarking infrastructure needed to systematically evaluate the robustness of emerging neural routing and data-driven approaches.

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.

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

TL;DR

SynthCharge is introduced, a parametric generator that produces diverse, feasibility-screened EVRPTW instances across varying spatiotemporal configurations and scalable customer counts, and provides the dynamic benchmarking infrastructure needed to systematically evaluate the robustness of emerging neural routing and data-driven approaches.

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.
Paper Structure (22 sections, 1 equation, 3 figures, 2 tables, 1 algorithm)

This paper contains 22 sections, 1 equation, 3 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. EVRPTW Formulations and Benchmark Suites
  4. Learning-Based Routing under Distribution Shift
  5. Synthetic Benchmarks and Validation Practices in Intelligent Transportation Systems
  6. Methodology
  7. Notation and Variable Definitions
  8. Spatial Topology Generation
  9. Charging Infrastructure Generation
  10. Vehicle and Service Parameters
  11. Time Window Assignment
  12. Feasibility Screening Procedure
  13. Stage 1: Structural Feasibility Screening
  14. Stage 2: Exact MILP Verification (Small Instances)
  15. Generation Algorithm
  16. ...and 7 more sections

Figures (3)

  • Figure 1: Representative SynthCharge spatial topologies under three spatial regimes: (a) random (R), (b) clustered (C), and (c) mixed (RC). The depot is shown as a square, customers as circles, and charging stations as triangles.
  • Figure 2: Acceptance rate $\gamma$ (%) versus problem size $N$ across spatiotemporal regimes.
  • Figure 3: Problem size versus mean traveled distance and mean EV count for random (R), clustered (C), and mixed (RC) regimes under a baseline metaheuristic. The trends illustrate how temporal tightening increases fleet requirements.