Case study
Bilbao Metro – Passenger Demand Prediction Model
Bilbao Metro is committed to advanced analytics and artificial intelligence to optimize the planning and operation of its network, through the design, validation and adjustment of a passenger demand prediction model developed by Virtual Desk.
In its commitment to innovation and continuous service improvement, Metro Bilbao needed a predictive model that could anticipate passenger demand based on multiple data sources: ticketing, sensors, open data, and contextual variables.
The main objective was to have a tool that would facilitate strategic planning, operational decision-making, and resource allocation based on projected passenger flow.
The immediate resolution of any incidents that occur on a daily basis is critical for the proper functioning of the service.
+ 5M €
In annual costs
90%
Prolonged delays
100%
Lack of interoperability
It is based on a big data platform that characterizes multimodal mobility in cities and builds multimodal chains.
- Large volume of scattered data: from ticketing, sensors, GIS and mobility surveys.
- Lack of its own Big Data infrastructure, which limited internal analytical capacity.
- Need for accurate forecasts that are adaptable to changes in habits and external events.
- Integration of multiple heterogeneous sources (ticketing, AEMET, economic variables).
- Robust statistical validation, which would guarantee the reliability of the model before its implementation in production.
Virtual Desk implemented an advanced solution based on its VDDP (Virtual Desk Data Platform), which enabled the ingestion, processing, and modeling of large volumes of information. The development included:
- Data quality analysis of the ticketing and validation system of Metro de Bilbao.
- Integration of external sources, such as meteorology (AEMET), events and economic variables.
- Design of multivariable predictive algorithms with continuous learning and adjustment capabilities.
- Construction of a prediction model in different dimensions: station, line, type of day, time slot and user profile.
- Statistical validation and quality control, guaranteeing an accuracy level of 80–85%.
- Final deployment of the analytical model and secure access for the Bilbao Metro technical team.
The project was developed under the Scrum methodology, with iterative cycles of development, review and validation, ensuring the delivery of partial results in each sprint.
This project has marked a milestone in our ability to understand and anticipate our users' mobility patterns. Thanks to the predictive model developed by Virtual Desk, we can plan more accurately, optimize resources, and improve the traveler experience.
Our platform guarantees accurate, real-time data exchange between institutions, improving reliability and providing a more efficient service for dependent individuals.
Ready to Transform Your Processes?
Cutting-edge Interoperability Solutions
Discover how Virtual Desk can optimize the systems of your public or private organization with advanced technology tailored to your specific needs.
Our solutions comply with the National Interoperability Framework (ENI) and are designed to integrate seamlessly with existing systems.
More case studies
Explore how we’ve helped other organizations optimize their processes and improve operational efficiency. Our personalized, results-oriented approach has demonstrated a significant impact across various sectors.
We design customized solutions that optimize processes, automate tasks, and free up resources. This translates into savings, increased productivity, and greater team satisfaction.