Ignite & Oxford launch AI tool for AV safety compliance

Ignite by FORVIA HELLA and Oxford Semantic Technologies have partnered to develop an explainable artificial intelligence tool for autonomous vehicle safety compliance. The software is designed to help manufacturers show how automated driving systems make decisions.

The project centres on simulation software built with RDFox, a knowledge graph database developed by Oxford Semantic Technologies, which is owned by Samsung Electronics. The system is intended to link traffic laws with live autonomous driving decisions and create a record that engineers and regulators can review.

The collaboration targets a persistent problem in the autonomous vehicle sector: technical progress in driving systems has outpaced the industry's ability to prove that those systems behave safely and legally across a wide range of road conditions. The issue has become more pressing as carmakers try to move beyond Level 2 driver assistance towards Level 3 and Level 4 automation, where responsibility shifts more heavily to the manufacturer.

Reasoning layer

Under current definitions, Level 2 systems can assist with steering, braking and acceleration, but the driver remains legally responsible for the vehicle. At Level 3, the system can take over certain driving tasks in defined conditions, while Level 4 covers full self-driving within specific operational domains.

The software adds a reasoning layer to autonomous vehicle systems by combining data with formalised expert knowledge and rules. In practice, that means translating traffic laws into machine-readable instructions, then checking vehicle behaviour against those rules in simulation and operation.

The approach is based on knowledge-based AI rather than machine learning alone. While machine learning systems infer patterns from large datasets, knowledge-based methods rely on structured rules and logic that can be inspected more directly when a vehicle makes a decision.

That distinction matters for safety cases and regulatory approval because it allows developers to trace an action back to the rule set that produced it. The companies argue this gives engineers a clearer view into systems often criticised as black boxes.

Safety scrutiny

The launch comes amid wider scrutiny of automated driving systems. The sector continues to face questions about how vehicles respond in unusual or fast-changing situations, including road hazards, weather events and shifting traffic conditions.

Recent incidents have kept those concerns in focus, including recalls involving self-driving vehicles in the United States after reports that some cars drove into flooded roads. Such cases have highlighted the difficulty of handling edge cases and showing that a system's decision-making meets legal and safety standards.

Ignite by FORVIA HELLA said the new system could reduce the need for manual, market-by-market coding of road rules, a process that can slow deployment as manufacturers adapt vehicles for different jurisdictions. It also said the software could produce evidence for European type approval by showing deterministic links between legal rules and vehicle behaviour.

"Hella Ignite.Drive applies knowledge-based AI by translating traffic laws, originally written for human interpretation, into machine-readable rule sets. This enables manufacturers to generate deterministic evidence that demonstrates safe and compliant vehicle behaviour for European type approval. At the same time, it reduces development lead times by minimising the need for manual, market-by-market rule coding, helping AV teams move faster toward approval-ready deployment," said Dr Felix Kortmann, Chief Technology Officer, Ignite by FORVIA HELLA.

Research origins

Oxford Semantic Technologies traces its roots to computer science research at the University of Oxford and was acquired by Samsung in 2024. RDFox is already used across sectors including financial services, manufacturing, healthcare, publishing and retail, where organisations need to combine large volumes of data with rule-based analysis.

In automotive, proponents of that approach say a vehicle can use knowledge-based AI as both a rulebook and a memory. By recording what the vehicle did, comparing it with applicable road rules and preserving the logic behind the outcome, developers can produce a more auditable account of vehicle behaviour.

"Autonomous vehicles currently use AI to make a whole range of decisions on the road, but at the moment manufacturers are struggling to show why or how these decisions are made. RDFox can help with this major barrier to progress. What knowledge-based AI allows us to do is to collect and map these decisions and apply reasoning. We can see exactly why a vehicle acts in a certain way and use this data to help the vehicle make better decisions in the future," said Peter Crocker, Chief Executive Officer, Oxford Semantic Technologies.

The work also reflects a broader shift in the automotive industry towards software-defined vehicles, where more of a car's functions are governed by software updates, digital architectures and centralised computing. As that shift continues, proving that software-driven decisions comply with road rules is becoming as important as the performance of the driving system itself.

"The AV case study is a great example of how knowledge-based AI can enhance data-driven systems. A key advantage of the technology is traceability, where decisions can be linked back to the rules and logic that produced them. In the automotive space, this visibility can revolutionise go-to-market strategies, improving the compliance and safety of AVs," added Dr Ian Horrocks, Oxford University Professor and Co-Founder, Oxford Semantic Technologies.