Network and Configuration Data for A Scheduling-Based Approach to Railway Capacity Estimation for Industrial Freight Junctions

This repository contains the infrastructure configuration files and scenario parameter files used in the experimental evaluation reported in the paper:

L. Di Gaspero, A. Grosso, G. Longo. A Scheduling-Based Approach to Railway Capacity Estimation for Industrial Freight Junctions. 32nd International Conference on Principles and Practice of Constraint Programming (CP 2026), Lisbon, Portugal, July 2026.

Overview

The dataset supports a Constraint Programming model for estimating the maximum freight throughput of a railway junction serving an industrial area with multiple terminals. The model is validated on a representative sub-network of a maritime port in the northern Adriatic Sea, comprising two operational buffers and four freight terminals.

Two infrastructure configurations are provided (as-is and to-be), each paired with a set of scenario configuration files that define the minimum and maximum number of freight services per terminal.

File Description

Infrastructure files (JSON)

These files define the physical railway network: endpoints, turnouts, buffers, terminals, track connections, and the passenger train timetable. All distances are in kilometres and all speeds in km/h; time quantities derived from these parameters are expressed in minutes.

Trieste2buff_asis.json — Current (as-is) infrastructure configuration.

Buffer b1: capacity 12 trains

Buffer b2: capacity 2 trains

Terminal capacities: T1=2, T2=4, T3=2, T4=2

Trieste2buff_tobe.json — Planned (to-be) infrastructure configuration after buffer expansion.

Buffer b1: capacity 20 trains

Buffer b2: capacity 10 trains

Terminal capacities: unchanged

Both files share the same network topology, track connections, and passenger train timetable (61 arrivals and 63 departures over a 24-hour period).

Scenario parameter files (YAML)

Each YAML file defines, for every terminal, the minimum number of mandatory freight services (min_service) and the upper bound on generated candidate services (max_service). When min_service == max_service, the solver operates in satisfiability mode; when min_service < max_service, it operates in maximisation mode.

The five scenarios reported in the paper are:

File
Scenario
Infrastructure
Mode

Trieste2buff_asis_solo_attuale.yaml
S1 — Baseline
as-is
satisfiability

Trieste2buff_asis_target.yaml
S2 — Authority target
as-is
satisfiability

Trieste2buff_asis_min_hb.yaml
S3 — Maximisation
as-is
maximisation

Trieste2buff_tobe_target.yaml
S4 — Authority target
to-be
satisfiability

Trieste2buff_tobe_min_hb.yaml
S5 — Maximisation
to-be
maximisation

Two additional scenario files are included as supplementary configurations:

File
Description

Trieste2buff_asis_min.yaml
As-is maximisation with conservative service upper bounds

Trieste2buff_tobe_min.yaml
To-be maximisation with conservative service upper bounds

Scenario details

S1 — As-Is Baseline (Satisfiability). Minimum and maximum service counts equal the number of trains currently operated at each terminal (T1: 3, T2: 6, T3: 1, T4: 1, total 11). Used as a correctness validation of the CP model against known operational data.

S2 — As-Is, Authority Target (Satisfiability). The regulatory authority has set a target of 20 freight services per day (T1: 3, T2: 11, T3: 5, T4: 1) on the as-is infrastructure. Assesses whether this target is achievable without any infrastructure investment.

S3 — As-Is, Maximisation. Minimum service floor equals current traffic levels; upper bounds are set well above any plausible infrastructure limit (T1: 12, T2: 24, T3: 6, T4: 6) to avoid artificially capping the result. Quantifies the true residual capacity of the existing infrastructure.

S4 — To-Be, Authority Target (Satisfiability). Planned infrastructure upgrade modelled with expanded buffer capacities. Minimum and maximum service counts equal the authority's long-term target (T1: 6, T2: 16, T3: 12, T4: 2, total 36). Verifies whether the planned upgrade is sufficient to meet future service objectives.

S5 — To-Be, Maximisation. Same to-be infrastructure. Conservative minimum service floor (T1: 3, T2: 6, T3: 3, T4: 1); upper bounds set well above any expected limit (T1: 12, T2: 24, T3: 12, T4: 6). Quantifies the capacity gain from the infrastructure improvement.

Usage

The configuration files are intended for use with the CP model implementation described in the paper, available in the accompanying code repository. The infrastructure JSON file is passed as the network description; the YAML file is passed as the scenario configuration. All experiments were run with OptalCP using 76 parallel workers and a time limit of 3600 seconds for maximisation scenarios and 600 seconds for satisfiability scenarios.

License

This dataset is released under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence.

Citation

If you use this dataset, please cite the associated paper:

@inproceedings{digaspero2026scheduling,
author = {Luca {Di Gaspero} and Alessia Grosso and Giovanni Longo},
title = {A Scheduling-Based Approach to Railway Capacity Estimation
for Industrial Freight Junctions},
booktitle = {32nd International Conference on Principles and Practice of
Constraint Programming (CP 2026)},
series = {Leibniz International Proceedings in Informatics (LIPIcs)},
volume = {379},
pages = {62:1--62:XX},
year = {2026},
publisher = {Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
doi = {10.4230/LIPIcs.CP.2026.62}
}