EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

2053. Using Constraint Programming to Plan Annual Maintenance of Railway Tracks

Invited abstract in session MD-56: Optimization in transportation infrastructure design and management, stream Transportation.

Monday, 14:30-16:00
Room: S04 (building: 101)

Authors (first author is the speaker)

1. Maria Andreina Francisco Rodriguez
Uppsala University
2. Justin Pearson
Institutionen for IT, Uppsala universitet
3. Jerker Sundström
Infranord

Abstract

Railway infrastructure maintenance is essential for the continuous operation of train transportation; it is costly and can only be performed in limited time windows. Maintenance companies are contracted to fix segments of railway tracks that fall below a certain quality threshold. We will be concerned with tamping, which is the process of smoothing out track irregularities.

Due to the short time windows, maintenance companies must choose which segments will be tamped each day. The tamping machines have fast travel times but are slow to tamp with significant starting and stopping times. Therefore, tamping as many contiguous segments as possible is more efficient. Before the end of a shift, the machine must return to the depot.

We use constraint programming to create multi-daily maintenance plans (and re-plans) that maximise the total length of tamped segments. Our test data is based on real-life planning requirements with instances of continuous tracks of approximately 120 KM divided into more than 800 segments of varying sizes from 30 to 400 meters. Our results are promising, and we can quickly find near-optimal solutions using Google OR-Tools. Our model has several non-obvious improvements, such as symmetry breaking and over-approximations to the upper bound, significantly improving solving time.

Keywords

Status: accepted

Back to the list of papers