Related Experiment Video
Updated: Jul 9, 2026
10:46
A Method of Trigonometric Modelling of Seasonal Variation Demonstrated with Multiple Sclerosis Relapse Data
Published on: December 9, 2015
10.7K
The MetroPT dataset for predictive maintenance.
Bruno Veloso1,2,3, Rita P Ribeiro4,5, João Gama4,6
1University Portucalense, Porto, 4200-072, Portugal. bruno.m.veloso@inesctec.pt.
Scientific Data
|December 13, 2022
Summary
The MetroPT dataset offers valuable sensor and GPS data for developing machine learning models for public transport predictive maintenance and anomaly detection.
Area of Science:
- Transportation Engineering
- Data Science
- Machine Learning
Background:
- Public transportation systems require robust maintenance strategies to ensure operational reliability.
- Predictive maintenance models can enhance efficiency and reduce downtime in urban transit.
Purpose of the Study:
- Introduce the MetroPT dataset for machine learning-based predictive maintenance.
- Facilitate the development of online anomaly detection and failure prediction methods for metro systems.
Main Methods:
- Collected diverse data from an urban metro service in Porto, Portugal, in 2022.
- Integrated analog sensor signals (pressure, temperature, current), digital signals, and GPS data.
- Structured the dataset for ease of use in developing machine learning algorithms.
Main Results:
- The MetroPT dataset provides a comprehensive framework for analyzing metro system performance.
- The data includes real-world operational parameters crucial for anomaly detection.
- The dataset serves as a benchmark for evaluating predictive maintenance models.
Conclusions:
- The MetroPT dataset is a valuable resource for advancing research in transportation maintenance.
- It supports the creation of advanced machine learning models for operational anomaly detection.
- This dataset can significantly contribute to improving the reliability of urban public transportation.
