From Crash Data to Safer Roads: in automotive and transportation industries, accident scenarios are the ultimate edge cases. They drive regulations, shape design requirements, and determine safety outcomes.
There are two big challenges:
- Connecting real crash data with standards (Euro NCAP, UNECE, ISO…)
- Recreating those scenarios automatically in simulation for testing & validation, used by not only requirement but also simulation, AI, control, V&V engineers,
Recently we have been working on an AI Agent app to deal with those obstacles, using Multi-modal Generative AI for:
🔹 Knowledge extraction from datasets like NHTSA crash reports
🔹 Mapping to standards with clear description, test criteria, parameters, and visualization
🔹 Interface and Automatic generation of simulation scenarios, i.e in ASAM e.V. OpenSCENARIO, ready to run in most industry tools (i.e Simcenter Prescan)
We exploit state of art GenAI technology to bridge the gap between real-world data and standards, accelerating multiple business needs:
✅ Insurances: faster, evidence-based decisions
✅ OEMs & Tiers: validate designs and safety compliance
✅ Legal teams & regulators: enhance standards
✅ Fleet operators: proactive training scenarios
Those are real values for reducing liability, accelerating validation, and improving safety at scale.