Euro NCAP car crash safety ratings get major four-pronged shake-up

Auto Express

3 days ago

A comprehensive overhaul of the Euro NCAP safety ratings will see current assessments replaced next year by ratings across 'four pillars of safety' – safe driving, crash avoidance, crash protection, and post-crash safety – from next year.
Euro NCAP will realign many of its existing test protocols into the four new categories, but there'll be new protocols too, especially in the area of ratings for Advanced Driver Assistance Systems (ADAS).
We caught up with Euro NCAP's Technical Manager ADAS/AD, Adriano Palao, during testing of the Renault 5 at MIRA, and he told us why the changes are needed. Advertisement - Article continues below
'We are very careful to only introduce scenarios in the assessment that can happen in real life,' he said, 'but there are missing pieces and we still have room for improvement.
'We started robustness activity around two and a half years ago, as we were keen to understand how sensitive our testing was to small changes. What happens when you move the car a little bit to the left or to the right? What happens in a scenario that was supposed to be completely free of objects, if you suddenly introduce a car next to the main target? These little variations actually have a tremendous effect on the performance of ADAS,' he claims. Skip advert Advertisement - Article continues below
'That led us to understand that some vehicle manufacturers were actually sub-optimising their systems to perform well in our tests, and not in real-world situations,' Palao told us, while also acknowledging others have 'a consistent and thorough approach' that optimises for the real world, and not for NCAP scenarios.
'As with every game that has a set of rules, you can find a trick to beat it,' he says. 'In our case there is a formula to get five stars and everyone knows how to do it. So from 2026, the way we are going to assess ADAS is going to change. We are going to reserve the right to inject layers of variation into test scenarios, to understand how sensitive systems are to changes. Because it is not our purpose to assess cars just for show, or just to say, 'Okay, this has five stars'. Our driving force has always been real-world safety outcomes.'
Euro NCAP will be updating its car safety testing in four key areas for 2026.
NCAP is strengthening its assessments of driver-monitoring systems to improve attention and driver engagement. Whereas drowsiness and distraction alerts are worth two points currently, from 2026 there'll be 25 points on offer for more advanced systems monitoring the driver's performance, with top ratings requiring real-time eye and head tracking, and not just sensors that can tell if you're gripping the steering wheel. Advertisement - Article continues below Skip advert Advertisement - Article continues below
There'll also be points awarded for systems that can recognise and respond to signs of drug or alcohol impairment, and bring a vehicle to a halt if a driver is unresponsive.
Euro NCAP is also introducing ratings for the human-machine interface, looking at the position, feel and accessibility of essential controls,
This category will include the more intensive approach to testing collision-avoidance systems such as Steering Assist and Automatic Emergency Braking – building on the work Euro NCAP has already been performing for more than a decade, by introducing tests that better represent real-world accident conditions.
Euro NCAP will build on its crash testing programme that focuses on the passive safety aspects of a car's design and specification. There'll be a wider focus on drivers of different shapes and sizes, to improve protection for older drivers and child occupants. There'll also be greater assessment of the risks in rollover situations, and closer focus on pedestrian safety, including structures around the windscreen.
The range of upgrades to the test protocol here covers systems like eCall, which in future must include information on the number of occupants in a car in automated emergency messages. This requires the detection of occupants even when the seatbelts are unbuckled. Advertisement - Article continues below Skip advert Advertisement - Article continues below
There are also updated ratings for occupant-extraction procedures, for example around opening doors if there has been a vehicle power failure.
There will also be new ratings for an electric car's ability to detect and manage fire risk.
New safety technologies aren't just a challenge that car manufacturers must get right. Euro NCAP itself must find ways of accurately rating the performance of vehicle systems designed to meet the increasingly tough criteria it sets. Which is why we find ourselves at the MIRA test track in Warwickshire on a bright spring morning in the company of the dedicated team charged with rating the performance of the all-new electric Renault 5.
We're there at the invitation of Euro NCAP itself, which is keen to communicate how it goes about testing Advanced Driver Assistance Systems (ADAS) such as Steering Assist and Collision avoidance, as part of an effort to promote driver understanding of this increasingly ubiquitous tech. So we're going to witness the organisation's latest Assisted Driving test protocol being carried out for the first time in the UK. Or at least we'll be witnessing some of it. Euro NCAP's full 2024 Assisted Driving Test and Assessment protocol runs to nearly 60 pages, and awards points for the performance during live testing of speed assistance, adaptive cruise control, steering assistance and collision avoidance. Advertisement - Article continues below Skip advert Advertisement - Article continues below
It doesn't end there either, because points are also awarded for how well manufacturers explain their systems to drivers through marketing and handbook content, and how drivers are able to engage with the systems in practice. Of course this is on top of the raft of passive crash and safety system tests carried out by Euro NCAP, which have all helped to make the organisation such an indisputably powerful force for change.
In order to rate one car's performance against another, Euro NCAP must be able to systematically repeat tests within an exact set of parameters. That's not easy when a number of the protocols include the vehicle being tested in scenarios including other vehicles – cars, motorcycles and even bicycles – which can either be moving or stationary.
We're here to see how NCAP assesses the performance of emergency brake assist when confronted by the particular challenge of a stationary or slow-moving obstruction that's revealed scant seconds before a potential accident, when when the car that you're following using active cruise control swerves sideways into the next lane to avoid its own collision. NCAP calls this a 'cut-out test' and awards points for active braking systems that will successfully prevent you from ploughing into the back of the stranded or slow-moving car, and more points if your emergency braking system can also avoid the motorcycle and bicycle – which are much more difficult for a test car's sensors to recognise and respond to. Advertisement - Article continues below Skip advert Advertisement - Article continues below
James Buck, technical specialist for the ADAS team at MIRA, introduces us to some of the test kit, which includes an electric Peugeot E-308 adorned with rooftop aerials and an interior stuffed with very expensive electronics, including state-of-the-art positioning technology and robotic controls in the form of a servo-operated steering wheel and pedals.
Inside the middle of the car, a solid vertical metal pillar supports an important-looking red box. 'That box needs to be rigidly attached to the vehicle so it doesn't judder around, because within it are accelerometers and gyroscopes, which work with the GPS antenna on top to give us an extremely precise location,' says James. 'Also on the roof is a WiFi antenna, so we can send our data over to a WiFi antenna on the test car, which has its own red box doing exactly the same thing, so both cars know exactly where the other is.
'The reason we've got the steering robot in is to make sure that when we do the test, it's very repeatable,' he continues. 'So we can do the test a hundred times and we can eliminate all variables in it. If I was to do the lane change myself around the bike, I'm never going to get it exactly the same every time. The steering robot is receiving information from the red boxes, so knows exactly where to go, and by having steering that is consistent and repeatable, we can isolate variables to make sure we're testing the vehicle's actual performance.' Advertisement - Article continues below Skip advert Advertisement - Article continues below
The Peugeot is referred to in the protocol as the 'SOV' or 'secondary other vehicle' in the cut-out tests, but the potential 'target' vehicles that the test car must try to avoid colliding with are pretty special too. Not at first sight, it should probably be said, because the Global Vehicle Target (GVT), Euro NCAP Motorcyclist Target (EMT) and Euro NCAP Bicyclist Target (EBTa), are outwardly rather flimsy-looking foam constructions with a decidedly Heath-Robinson flavour.
There's a good reason for that, of course, because if a collision-avoidance system fails to avoid a collision on test, the test car and occupants must be able to drive straight through the target, sending foam components flying, but leaving the test car undamaged.
'A lot of effort goes into making the targets visually correct, so the camera system thinks it's real, but the radar characteristics have to be right as well, because obviously they have to reflect like a real-life target to be picked up by vehicle radar and LiDAR [(Light Detection and Ranging] systems as well,' says James. 'So as well as the foam, they contain metallic plates or shavings hidden deep inside, so to a test vehicle's sensors they'll look just like the real thing.'
Of course the targets also have to be moving for the Euro NCAP tests, which means they scoot around on computer-controlled trolleys which are just a few centimetres off the ground, allowing them to pass safely underneath a car's wheels if a collision takes place. 'The Global Vehicle Target is like a big flying carpet with a foam body on top,' James quips. 'But again, it's had a lot of development work to make sure it looks like a real car to the radar sensors.' Advertisement - Article continues below Skip advert Advertisement - Article continues below
It's certainly reassuring to know that we're dealing with the 'soft target' motorcycle when it's our turn to jump in the Renault 5 and experience the test for ourselves.
Interestingly, while the SOV is driven by robots, there's a test engineer in the driving seat of the Renault. Ameel Lalji leads the Euro NCAP ADAS programme at MIRA, and as we set up for our first collision-avoidance 'cut-out test', he explains the process.
'The car is going to pretty much do its own thing,' he tells us. 'I'm going to activate the Adaptive Cruise Control [ACC], set it to the closest gap to the vehicle ahead, and a speed of 70kmh. The SOV will be driving at 50mph, so we'll close in and let our car hold position. Then I'll let go of the wheel, as the car will be doing its own steering to stay in the lane.'
It's getting quite exciting now, especially when Ameel goes on to explain what happens next: 'I should warn you the braking is going to be a bit aggressive,' he says. 'Because we're following under ACC, when the SOV ahead of us cuts out of our lane, the first thing our car will do is try to accelerate up to its set speed of 70kmh. Then, when it recognises the obstruction ahead, it should slam on emergency braking.' Advertisement - Article continues below Skip advert Advertisement - Article continues below
Luckily the Renault performs exactly as hoped, and our foam motorcyclist survives the experience, braking sharply to a halt five metres short of an impact. Ameel checks his laptop and reads out the numbers: 'That was nine metres per second of deceleration, so pretty much full on at almost 1g of braking. It won't do much more than that, but it only had to do it for a second and a half,' he says.
So could – or should – a real-life driver on the motorway have performed as well as the ADAS system in a real-life emergency? Well maybe, but what if they were momentarily distracted by their touchscreen or chatting to a passenger? In that case the motorcyclist might not have fared so well, and from our point of view the test was extremely affirming.
As driving enthusiasts, we may not always appreciate annoying interventions from ADAS going about its business as we're going about ours, but do we really need to have to deal with the consequences of cars striking bikers or cyclists in real life to appreciate their value? We'd say no, but it's clear from our own experiences of ADAS and those related by readers, that more development is needed in some cases to ensure blemish-free operation. But intense pressure to improve standards is exactly what the Euro NCAP programme has always delivered and, unlike our Renault, it's not slowing down for the obstacles ahead.
Click here for our list of the best city cars on sale ...
Find a car with the experts Car Deal of the Day: Super sci-fi Kia EV6 for an exceptional £260 per month
Car Deal of the Day: Super sci-fi Kia EV6 for an exceptional £260 per month
Our Deal of the Day for 31 May is an extremely attractive price for this extremely capable electric family car Car Deal of the Day: new Omoda 5, with all the kit you might need, for less than £200 per month
Car Deal of the Day: new Omoda 5, with all the kit you might need, for less than £200 per month
The Omoda 5 is one of the newest arrivals to the family SUV market, and now it's our Deal of the Day for 1 June Audi A3 Sportback 35 TFSI long term test: premium hatch is a joy to live with
Audi A3 Sportback 35 TFSI long term test: premium hatch is a joy to live with
First report: premium hatch joins our fleet, and its bright metallic paint has already earned it a nickname Long-term tests
1 Jun 2025