BEFAHRBAR – Methodology for Evaluating Potential Areas of Application for Autonomous Vehicles in Terms of Safe, Efficient Drivability
The BEFAHRBAR project is developing an automated methodology based on a digital twin to safely and efficiently evaluate routes for the deployment of autonomous SAE L4 vehicles, while also objectively identifying risks. Salzburg Research is providing the key scientific foundations for this, as well as playing a major role in designing a transparent and practical evaluation and approval process for automated driving.
Automated mobility is a vital part of the transition to a new approach to mobility. It is essential for achieving the goals of the Mobility Master Plan 2030, such as improving traffic safety and efficiency and promoting climate neutrality. However, the widespread use of SAE Level 4 vehicles in public transport presents significant challenges, primarily due to the technical, organizational, and legal requirements involved. The complex approval, risk analysis, and mitigation processes required to ensure safe operation, in particular, are an obstacle to scalable deployment. Transitioning from pilot operations to real-world operations requires innovative testing and validation methods that consider national framework conditions, manufacturer requirements, and the specific requirements of automated vehicles uniformly.
The aim of ‘BEFAHRBAR’ is to develop a methodology for the automated evaluation, monitoring, and analysis of SAE L4 vehicles in public transport that can be scaled up. To this end, the most efficient methods possible are to be used to enable a well-founded analysis and evaluation of the safe navigability of the route. A highly accurate, simulation-capable model of the route based on a digital twin will make route and risk assessments and make development and monitoring processes more efficient, robust, objective, and adaptable. This will not only facilitate compliance with legal requirements but also address operational challenges and technical conditions.
To this end, a realistic, simulation-capable model of the route is being developed to enable efficient and robust analysis and evaluation of safe drivability. This is based on a highly accurate digital twin of the route’s spatial data, created using mobile mapping. This is being expanded to include methods for the automated evaluation of the environment and the simulation of dynamic environmental elements, such as weather, traffic, and different SAE L4 vehicles.
The simulation-capable model performs the following tasks:
- Recording and documentation of relevant environmental conditions
- Comparison of relevant and specific environmental conditions with
o National legal framework
o Operational requirements and
o Vehicle-specific ODD - Identification of objective risk factors for individual sections and the entire route
- Derivation, validation, and monitoring of measures to ensure operational safety (ODD requirements and mitigation of identified risks)
- Recording and analysis of critical situations during operation
The developed methods, models, and procedures are validated and improved using experience gained from automated vehicle test operations, existing route and risk assessments, and test drives at existing test sites. The validated methodology is then applied as a proof of concept on a representative route, taking into account the operator’s real-world requirements.
Ultimately, these methods should be used to make the approval, routing, and risk assessment processes as efficient and well-founded as possible, while also supporting deployment measures and ongoing operational and risk monitoring.
Role of Salzburg Research
Together with its project partners, Salzburg Research is analyzing the existing evaluation scheme for test applications for the approval of automated driving on defined test tracks. Based on the results of this analysis, the company is supporting the development of a digitized evaluation approach based on a highly accurate digital twin.
Salzburg Research is taking on the central leadership role in Work Package 4, which establishes the scientific basis for the partially automated assessment of risks associated with the use of automated vehicles. To this end, a method is being developed to divide the ‘digital twin’—a digital representation of the real environment—into meaningful areas for evaluation in accordance with the respective context. Additionally, the project will examine the most effective measures to promote the safe use of automated vehicles.
The developed methods enable individual sections of a route to be assessed objectively and comprehensively. By subsequently evaluating and validating the methods using real data, Salzburg Research is making a significant contribution to developing a transparent, objective, and practical approval process for automated driving in Austria.
