Simulated and Real-World Urban Mobility Trajectories from Barcelona

Two datasets from the CODECO European research project focus on urban road safety. The first contains simulated vehicle trajectories for collision scenarios in Barcelona, while the second includes real-world trajectories of Vulnerable Road Users and vehicles from the UPC campus and i2CAT offices. Both are intended for training machine learning algorithms for collision prediction and risk detection.

Use Cases

• Train ML-based collision prediction algorithms using simulated vehicle trajectory records designed to provoke lateral and frontal collisions.
• Improve Vulnerable Road User (VRU) risk detection algorithms with real-world trajectory and mobility records of pedestrians, cyclists, and scooter riders.
• Evaluate vehicle-to-everything (V2X) communication and orchestration models using data fields aligned with ETSI ITS VAM and CAM message standards.
• Analyze reckless driving behaviors such as traffic light violations and speeding within simulated urban scenarios geolocated in Barcelona.

Strengths

• Data is produced within a formal European research project (Horizon Europe, Grant Agreement No. 101092696).
• Includes both simulated scenarios for controlled testing and real-world observations for validation.
• Real-world data covers multiple types of road users (VRUs and motor vehicles) collected over multiple days.
• Released under the ODC Open Database License (ODbL) with clear project documentation.

Limitations

• The exact scale (row count, file size) is unspecified, limiting assessment of dataset scope for model training.
• Simulated data may not fully capture the complexity and noise of real-world traffic environments.
• Real-world data collection is geographically limited to specific sites (UPC campus, i2CAT offices) in Barcelona.

Provenance

Source

CODECO (COgnitive DECentralized Orchestration) European research project.

Collection Method

HEU-101092696-CODECO-Trajectories: Generated using the SUMO traffic simulator with TraCI interface. HEU-101092696-CODECO-VRUMobility: Collected using camera sensors and V2X communications.

Time Range

null

Freshness

null

Geography

Central Barcelona (simulated data); UPC campus and i2CAT offices in Barcelona (real-world data).