Are ADAS applications making us worse drivers?
I was struck by something recently. Most cars on the road today are filled with ADAS. Advanced Driver-Assistance Systems paint a pretty picture of an idyllic world in which all vehicles drive along in harmony, rainbows fluttering out the exhaust and unicorns prancing alongside.
The technology is certainly evolving, and quickly. It’s impressive to see, likely leading to fewer collisions despite its far-from-perfect applications.
But is it really helping us as humans?
Well, perhaps sometimes, but as it stands at the moment, I don’t think so. At least, the majority of ADAS applications are making problems of their own.
Here’s why.
ADAS and my automatic high beam headlights
My car, a Hyundai Santa Fe, has a few ADAS applications (but less than most!). The most apparent to me is the automatic high beam.
Theoretically, it works as follows. Twist the headlights from OFF to AUTO (see the image above). When you push the stalk back, the automatic high-beam system activates.
Full-beam headlights will turn on if the car’s camera registers that no other vehicles are in front of you or driving towards you in the other lanes. If it detects these, it switches the full beam lights off, leaving you with the standard dipped beam. Once no one’s at risk of getting dazzled, they turn back on.
Great.
Note that there are other versions of the automatic high beam system. Volkswagen’s, for example, is a little more advanced.
My question is — do we need it? What’s the point?
The supposed point
Now, you might argue that it prevents other bad drivers from dazzling you as you drive, and I agree. That’s great.
But the problem isn’t solved. It doesn’t address the bad driver but covers up their actions. One might argue it actually gives someone an excuse to not bother or worry about the high beam at all!
And when did lightly tapping a stalk — which, by the way, is an inch or two from my left hand on the steering wheel anyway — to manually turn off the high beams become such a challenging chore?
You’ll notice that, on the photograph of the stalk shown above, ‘AUTO’ (automatic headlight control) comes sequentially before the parking lights and normal headlights.
In other words, the manufacturer expects drivers to use ‘AUTO’ more than the manual controls. And they’re probably right.
That’s the real root of my thought here, which I’ll delve into now.
ADAS isn’t solving all our problems…
I’ll continue using the example of automatic headlights because it’s arguably one of the more straightforward ADAS examples.
Get Ben Kitchen’s stories in your inbox
Join Medium for free to get updates from this writer.
Despite the perceived benefits of automatic high-beam lights, I feel serious problems exist. Here are a few that spring to my mind:
- The ADAS high beams turn off when they see cars or headlights coming. However, that’s really too late. The other driver has already been dazzled, even if temporarily. A driver should have the awareness to manually dim the lights before the other car comes into view or range.
- Once I pass the other car, the full beams take a few seconds to turn back on, no doubt verifying that there’s no longer a vehicle there. What if something is happening up ahead that I need to see?
- The system doesn’t always work. I don’t know what you may have experienced in your daily driver, but I’d say one time out of seven, my full-beam lights stay switched on and dazzle oncoming cars anyway. They don’t always switch off when they should.
- Automatic high beams often mistake reflections in road signs for other cars and switch the beams off. Yes, technology will improve, and this will get better over time, but the fact remains that sometimes high beams switch off when you need them on.
This list highlights the elephant in the room regarding ADAS. Implementing a system that helps without completely usurping control is very challenging. In our headlight example, the only way to override the mistakes is to constantly switch the system back and forth to its manual mode — at this point, one may as well simply leave it there and trust their judgment.
Perhaps other drivers are getting dazzled less (that could be incorrect — reports suggest the problem is getting worse!), but we aren’t solving everything. In addressing the old issues, we make new ones for ourselves.
ADAS doesn’t always just help… it takes over
I’m not using this heading to suggest the cars will rise up and destroy us or anything like that! What I really mean is that giving the driver less responsibility makes for worse driving.
As you would expect.
In short, why do we depend on algorithms and software — which can always glitch or fail somehow! — to control the high beams? Why are we not instead implementing stricter penalties when drivers dazzle others? Are these systems really safer than simple driver training and awareness?
Why is the responsibility placed on the car’s system rather than the operator?
I don’t have access to studies or facts and figures (but I would leave to see any that readers know about!). However, I suspect we’re digging a bigger hole rather than finding the optimal solution.
I’m also sure that ADAS is a sort of halfway point as manufacturers seek to be the first to develop fully self-driving cars. And I think that’s why these systems — the ones that override rather than help the driver — aren’t that great.
In fairness, I can see the case for the more advanced ADAS systems, such as VW’s high-beam assist. These are far cleverer and usually work much better without compromising any driver’s vision. In no way do I think we should strip all modern cars of ADAS! It’s here to stay, and rightly so.
But there’s a fundamental underlying problem here, and one which no amount of ADAS will address. That is, quite simply, poor driving.
(In fact, as I’ve alluded to on this page, it might be argued that the cleverer ADAS becomes, the more it enables bad driving to go unchecked!)
How can ADAS be more effectively used?
As the name suggests, ADAS should assist the driver, not do the driver’s job for them.
We’ve seen an ever-increasing number of similar case studies worldwide in all sorts of fields and industries. Just because technology is improving doesn’t mean it overrides a person’s input.
Take content AI, for example. The marketing industry was turned upside down for a few months there last year. Would AI take over my job as a writer?
No. It’s a helpful tool for inspiration and structure, but I’m still here to write the content. It’s actually helping me, not hindering me.
Consider VAR in football (soccer), where its implementation has been taken… badly, to say the least. During its first couple of seasons, VAR would often override the on-pitch referee’s decision before he’d even made it, making his position appear vital but then be shown as unimportant and useless as the technology undermined him.
At the moment, many ADAS applications work like this. A driver is no different from that on-field ref.
Better-designed systems take technology and use it to improve the driver’s attention and reactions. These include warning systems, live OBD (on-board diagnostics) monitoring, and perhaps HUDs (heads-up displays).
Over the coming years, let’s hope technology helps us become better drivers, not better passengers.
