1825. Why Not Use Street Network Data for Visually Attractive Waste Collection Routes?
Invited abstract in session TD-58: Machine Learning and Artificial Intelligence, stream Vehicle Routing and Logistics.
Tuesday, 14:30-16:00Room: Liberty 1.13
Authors (first author is the speaker)
| 1. | Byung-In Kim
|
| Industrial and Management Engineering, Pohang University of Science and Technology (POSTECH) | |
| 2. | Seungyeop Lee
|
| Industrial and Management Engineering, Pohang University of Science and Technology (POSTECH) | |
| 3. | Sangil Han
|
| Industrial and Management Engineering, Pohang University of Science and Technology (POSTECH) |
Abstract
The residential waste collection problem (RWCP) is an arc routing problem where vehicles must service road segments under capacity and operational constraints. Traditional approaches focus on cost minimization, but factors like turn penalties, workload balance, and visual attractiveness (VA) impact feasibility. Street network data (SND) is essential for generating practical, visually appealing solutions, yet it has mainly been used for origin-destination (OD) matrix calculations rather than clustering and routing. This study leverages SND with a graph partitioning algorithm to create visually attractive clusters and integrates VA and turn penalties into an optimization framework. The arc overlapping index (AOI) quantifies resident inconvenience from odor, noise, and frequent truck visits. A multi-objective mixed-integer linear programming model optimizes routing within each cluster. The proposed matheuristic balances operational efficiency with practitioner acceptance. New RWCP benchmark instances, incorporating real-world data, validate the approach. Results show significant VA improvements with minimal cost increase, offering a scalable, practical solution.
Keywords
- Vehicle Routing
- Metaheuristics
- Mathematical Programming
Status: accepted
Back to the list of papers