A Practical Approach to Turning Research into Real-World Value

The BERTHA project introduces a robust good practice for transforming human driving behaviour research into reusable, validated and uptake-ready assets for the Connected, Cooperative and Automated Mobility (CCAM) ecosystem. This structured approach ensures that research outputs go beyond theory and become practical, reusable tools that support simulation, validation and innovation across the mobility sector.

From Data to Deployment: A Structured Methodology

One of the main challenges in CCAM research is bridging the gap between behavioural insights and engineering applications. BERTHA addresses this by providing a clear, step-by-step framework that guides the transformation of human driving data into deployable solutions.

1. Understanding Human Driving Behaviour

The process begins with the collection and analysis of real-world and experimental driving data. This enables the identification of key behavioural indicators, such as:

• Decision-making patterns
• Reaction times
• Risk perception

These insights provide the scientific basis for all subsequent developments.

2. Developing Probabilistic Driver Behaviour Models

Behavioural data is translated into probabilistic models that capture:

• Variability in human driving behaviour
• Uncertainty in decision-making
• Different driving styles and responses

These models are designed to be realistic, flexible and applicable to a wide range of CCAM use cases.

3. Making Models Accessible via an Open-Source HUB

To maximise reuse, BERTHA integrates its models into a centralised open-source HUB, ensuring:

• Easy access for external stakeholders
• Standardised and interoperable formats
• Compatibility with simulation environments

This step is essential to enable adoption beyond the project consortium.

4. Testing in CARLA Simulation Environments

The models are implemented and tested in CARLA-based simulations, allowing:

• Safe and controlled testing of driving scenarios
• Evaluation of interactions between human-driven and automated vehicles
• Reproducible and scalable experimentation

5. Validating through Demonstrators

The final step focuses on validation, using demonstrators and test cases to:

• Assess realism and performance
• Ensure reliability of the models
• Refine outputs through iterative improvements

This ensures that the developed assets are ready for practical use.

Designed for Reuse: Clear and Structured Outputs

A key strength of this good practice lies in the way BERTHA documents its results. Each asset is accompanied by clear and structured information, including:

• Purpose – what the asset is designed for
• Target users – who can benefit from it
• Validation evidence – how its reliability has been assessed
• Reuse conditions – technical and licensing requirements
• Exploitation routes – how it can be further used or developed

This makes BERTHA outputs not only accessible, but also easy to understand, test and integrate.

Who Can Benefit?

This good practice is designed to support a wide range of stakeholders, including:

• Researchers and academic institutions
• Simulation and software developers
• OEMs and Tier suppliers
• Safety and regulatory authorities
• Other CCAM and Horizon Europe projects

Supporting Impact Beyond the Project

By focusing on validation, accessibility and reuse, BERTHA contributes to the broader goals of Horizon Europe by:

• Enabling real uptake of research results
• Promoting collaboration and interoperability
• Supporting the development of safer automated mobility systems
• Creating long-term value for the CCAM ecosystem

A Replicable Approach for Future Projects

The BERTHA approach invites stakeholders to explore, test and build upon its outputs, fostering collaboration and accelerating innovation in automated mobility. By combining behavioural science, modelling, simulation and validation, this approach provides a replicable framework for future CCAM initiatives.