What we believe
Seatbelts were once controversial. So were airbags and electronic stability control. Each was doubted, and each is now so normal that nobody remembers the argument. Automated driving is the next step on the same road.
When a safer technology is held back, the people it would have saved never make the news. We count them anyway.
A death in a crash makes the news. A death that a better system would have prevented does not, because nobody can point to that person. That makes delay look free to the people who decide on it. Our paper The Cost of Waiting will estimate that loss per member state, with the method in public.
The fair benchmark for an automated system is the human driver on the same road: sometimes tired, sometimes distracted, sometimes drunk. A flawless driver does not exist.
Some people want automated systems to be flawless before they may drive. No human driver meets that standard, and we still hand out driving licences every day. The question that matters is whether a system causes fewer deaths and injuries than the drivers it replaces, on the same roads. Answering it needs good data on human drivers per country and road type. Our paper The European Human Benchmark sets out to collect it.
Set clear safety targets and measure the results in public. Leave the engineering to engineers.
Rules that prescribe how a system must work fix today's engineering in law. Rules that set a safety outcome leave engineers free to find better ways to reach it. Regulators should define what safe enough means and test it against data from real roads. When a system falls short, they should act quickly.
These systems improve with every update. Oversight should be continuous and based on data, so a safer version reaches drivers in weeks instead of years.
Type approval was designed for products that stay the same once they leave the factory. The software in a modern car changes after sale, often to fix faults or to make the car safer. When each change waits for a long approval cycle, drivers keep the older version for longer. Oversight should follow the data from the road, all the time.
Automated driving gives independence back to people who cannot drive: older people, people with disabilities, and people who live far from public transport.
Many people cannot drive because of age, a disability, a medical condition or where they live. For them, every trip depends on a timetable or on someone else's free time. A vehicle that drives itself could give them that control back. Our paper Mobility for All looks at who gains and which rules would need to change.
We want the world's safest cars designed, tested and built here, with European jobs and European standards.
Europe has the engineers, suppliers and car makers to lead in automated driving. If the safety case is proven only in other markets, the jobs and the standards will settle there too. We want European rules that are strict about outcomes and quick to apply, so the work happens here.
Progress does not happen by itself. It comes from people who believe something better is possible and then do the work to prove it.
Seatbelts, airbags and stability control did not arrive by themselves. Engineers built them, researchers measured them and campaigners argued for them, often for years. We see our work in that tradition. We gather the evidence and argue in public for rules that follow it.
We publish our funders and our data. Anyone who argues from evidence has to show their own.
An organisation that asks others to publish their safety data should publish its own finances. We list every funder who gives more than €1,000 a year. No single company provides more than half of our budget. Each paper comes with its data and method, so anyone can redo our sums and show us where we went wrong.