Data is a centrepiece of the Driver Behavioural Model (DBM) development in the BERTHA project, which needs comprehensive and reliable data to predict and respond to driver behaviours accurately.

However, in early 2024, it faced a significant challenge to its success: defining a unified data model and acquisition principles. This task has been essential to ensure all partners can collaborate effectively, share data, and produce consistent and reliable results across the various work packages.

Recognising the importance of data structure, BERTHA’s partner VEDECOM took the lead, focusing on defining the data collection and structuring methodology. The goal was to create a comprehensive data model encompassing the different data types, and guidelines for data acquisition, sharing, and formats.

Data collection common acquisition concerns, an important milestone for the BERTHA project

One of the main achievements of these efforts was the creation of the BERTHA project’s Data Model (Deliverable 2.1), which established the foundational data structures necessary for the DBM.

This was followed by the BERTHA Data Format & Acquisition Principles (Deliverable 2.2), which standardised the project’s data acquisition and sharing requirements and played an important role in achieving BERTHA’s Milestone 1, which involved identifying models influencing factors and addressing common concerns in data collection.

By providing a standardised framework, the unified data model will ensure that all partners can collect data effectively, facilitating integration and supporting the development of the DBM. The implementation of standardised processes and clear data-sharing agreements will maintain data integrity and quality throughout the BERTHA project.

Moreover, the unified data approach will create an interoperable data storage and exchange solution, which will be used for the development of the DBM during the experimental phase of the project in 2025.

A collaborative effort for the BERTHA project

As the leader of Work Package 2 on ‘Human driving performance data collection‘, VEDECOM coordinated the establishment of a unified data model and facilitated collaboration among all involved partners, including AIT, CVC, Continental Engineering, DFKI, Europcar Mobility Group, IBV, University Gustave Eiffel, and others.

A series of bi-weekly meetings was organised over the spring and summer of 2024. This collaborative effort was instrumental in reaching a consensus on the data structures and acquisition principles that would be acceptable and usable by all parties involved.

Only through their active participation was it possible to establish a comprehensive framework for data acquisition, format standardisation, and data structure, ensuring partners can collaborate effectively and achieve BERTHA’s scientific objectives.