FlexCrash develops an online simulation platform to enhance road safety by studying human and autonomous vehicle interactions 

DThe innovative FlexCrash platform, designed and deployed by the IMC University of Applied Sciences Krems, allows human drivers and autonomous vehicles (AVs) to interact in realistic simulations, offering insights into how these vehicles interface in varied and complex traffic environments, helping developers, regulators and traffic experts identify potential safety issues and optimize AV performance.
Current AV testing approaches strongly rely on simulations, but existing tools often fail to account for the complexities of mixed-traffic scenarios, involving both AVs and human drivers. FlexCrash fills this gap, by allowing integrations of multiple AVs and human drivers in a single simulation environment.

The platform adopts a turn-based strategic game paradigm, simplifying the simulation process and reducing cognitive load on users. This innovative approach also ensures that simulations can run smoothly even with network interruptions or delays.
The FlexCrash platform follows an established architecture design for scalable and user-friendly web platforms. It exposes a set of REST APIs and WEB GUIs for remote participations of AVs and human drivers.

“As the use of AVs is growing and becoming more common, understanding these interactions will be crucial for enhancing road safety and optimising traffic systems”, says Hannah Arpke, project’s coordinator at Eurecat.

Key functionalities of the platform include the creation and monitoring of mixed-traffic scenarios, sampling vehicle trajectories, and analysing the evolution of these scenarios over time. Users can design custom scenarios, track their progress in real-time, and assess vehicle behaviours to improve both autonomous and human driving strategies. The platform ensures each simulation concludes under specific conditions, such as reaching maximum duration or when all vehicles have either achieved their goals or experienced a crash event. Additionally, the platform includes a unique mediated trajectory sampling feature, enabling human drivers to plan their trajectories interactively, without needing to master complex kinematics computations.

“Instead of relying only on past accidents, the FlexCrash platform will generate realistic traffic scenarios based on actual driver behaviour, helping to detect critical situations and accidents not covered by existing datasets and crash databases”, highlights Deepak Dhungana, the head of institute for digitalization, one of the project partners from IMC University of Applied Sciences Krems.
While the current version of FlexCrash platform is already a powerful tool for small-scale experiments, the research team plans to enhance its performance and scalability in future updates.

Beyond simulation, the FlexCrash project also aims to enhance vehicle safety and sustainability through advanced crash-tolerant structures. By leveraging high-strength green aluminium alloys, the project seeks to develop safer, lighter, and more circular vehicle structures. This approach not only reduces greenhouse gas emissions but also significantly improves crashworthiness, with a particular focus on frontal impacts, which account for 70% of all car collisions.

The FlexCrash project is funded by the European Union’s Horizon Europe programme and its consortium is made up of ten partners from Spain, Italy, Germany, Sweden and Austria. They include four research and technology organisations (the Eurecat technology centre, the Stellantis’s CRF research centre, the Virtual Vehicle Research GmbH and Fraunhofer IWS), two universities (Luleå Tekniska Universitet and IMC University of Applied Sciences Krems), three industrial partners (Gestamp, Gemmate Technologies and Aerobase Innovations), and one standardisation body (UNE – the Spanish Association for Standardisation).