Car safety technology problems
How do we safely introduce a barrage of safety technology from many different suppliers into cars that can travel at 110km/h?
It’s terrific that we have high automotive safety standards in Australia and features such as ABS and electronic stability control are mandatory in vehicles that are being sold to the public and registered to drive on the road. After all, these one to three tonne metal blocks can easily become out of control projectiles.
And so now we find ourselves at the step of the next phase: mandatory requirements for electronically aided safety mechanisms, with the industry at a point where standardisation and minimum requirements for things like automatic emergency braking (AEB), adaptive cruise control and lane-keeping assist – even alcohol detection devices – are soon becoming legal must-haves to sell a vehicle in certain markets.
For a start, Europe will introduce some of these requirements this decade, and Australia is likely to follow suit. And even if Australia lags behind on timing, production efficiencies might just see that whatever is standard in European cars will end up shipped here anyway.
While that is mostly a good thing, the sudden expectations on vehicles to be able to do everything from automatically brake in all conditions for objects such as cars, pedestrians and cyclists and more, and to drive itself back into lanes without driver input is not an exact science.
Take Volvo for example, one of the automotive leaders for vehicle safety, which displayed its pedestrian AEB technology to journalists in 2016 on a V60 travelling toward a dummy named Bob. The Volvo hit Bob. It should have stopped. And this was all in a controlled environment, mind.
And just a couple years ago an autonomous Uber fitted with more sensors than a normal road car killed a pedestrian at night in the US while conducting public testing.
It shows that while we think this technology is good enough to be rolled out, it doesn’t mean it will always work. And while drivers are the ones that must always be compos mentis at the wheel, the outcomes from these systems can sometimes be worse than simply failing to engage.
We mentioned in our Isuzu D-Max review that we found the AEB and lane-keeping assistance system to work quite well and not false trigger, and that led to a few questions – which ones don’t and why not?
We’ve actually had some pretty terrible moments over recent years with false-positives in vehicles from Mitsubishis to BMWs. I had a Mitsubishi Eclipse Cross and it would trigger AEB when it shouldn’t, which meant it would bang on the brakes unexpectedly on a clear road with traffic behind. This included on the freeway. And there was a BMW X3 with lane-keeping assist that would steer off the road on one particular corner and toward a pole near a freeway turn-off at another. The steering assist can be quite strong in most cars, too.
This raises obvious safety problems. Sure, they mostly work for the better, but with pressure on manufacturers to pack in everything possible and keep models at a reasonable price point, there’s room for lacklustre performance that can potentially become unsafe.
There’s also the point that while some drivers we speak to don’t even know half of this stuff exists in the car and don’t use it, there are some that rely on it too much. It’s all too easy to break the law and get on the phone now with things like adaptive cruise control and lane-keep assist providing a false sense of safety – we’ve seen images already from phone cameras in Australia with drivers watching videos. But if one of the safety systems fails and steers toward a pole, you’re in a hell of a lot of trouble. It might even steer you toward another car.
And this isn’t just confined to what we’ve experienced. Mazda last year recalled its new Mazda 3 because the AEB was found to unexpectedly and inadvertently engage while driving. We experienced this among other writers we know. In 2018, the USA’s Insurance Institute For Highway Safety (IIHS) took five new vehicles – the Tesla Model 3 and S, BMW 5 Series, Mercedes E-Class and Volvo S90 – and tested AEB, adaptive cruise control and lane-keeping assist. Some of the problems encountered were that AEB didn’t actually work in some vehicles in some circumstances.
In other tests, the IIHS noted: “The 5-series steered toward or across the lane line regularly, requiring drivers to override the steering support to get it back on track. Sometimes the car disengaged steering assistance on its own. The car failed to stay in the lane on all 14 valid trials. The Model S was errant in the hill tests.”
This opens a further can of worms, which becomes the fact that we shouldn’t rely heavily on this technology because it can fail to work. That’s something we don’t hear much about.
But back on point, it can also actively do the wrong thing like hit the brakes on the freeway or steer towards an object at speed. The outcome of what can happen is particularly precarious in situations where the driver might not know the system exists until it falsely triggers near traffic or an obstacle.
In previous conversations with ANCAP, a representative said AEB and lane-keeping assist technology, which is where the car will steer itself, was beginning to be put under the microscope for how accurate it is and if in fact it would do the opposite and steer in a dangerous manner. And perhaps this should indeed be a priority over just saying the technology is in the car and ticking that box.