By Milad Haghani, The University of Melbourne; Akshay Vij, University of South Australia; Ali Ardeshiri, University of South Australia, and Zahra Shahhoseini, Monash University*
Cars are getting smarter. Today’s vehicles can automatically brake to avoid a rear-end collision, keep themselves centred in a lane, warn of hazards in blind spots and even maintain a safe distance from the car ahead.
Collectively known as advanced driver-assistance systems (or ADAS), these features have been shown to reduce crashes, injuries and insurance claims.
But there’s a problem: many drivers don’t want them.
In Australia, one in five motorists with cars equipped with these features have switched at least one of them off. In some countries, the figure is much greater.
These systems help prevent some of the most common and costly crashes on our roads. So why do so many drivers go out of their way to turn them off?
Why advanced driver-assistance systems matter
Advanced driver-assistance systems are a range of in-car technologies that help prevent or mitigate crashes.
Some of the most common include:
-
autonomous emergency braking: brakes automatically if a collision is imminent
-
lane-keeping assist: nudges the steering to keep you within lane markings
-
blind-spot monitoring: alerts you if another vehicle is in your blind spot
-
adaptive cruise control: maintains a set speed while keeping a safe distance from the vehicle ahead.
These systems aren’t experimental. They have proven safety benefits.
Data from the United States shows autonomous emergency braking cuts rear-end collisions by 50%. Even forward-collision warning on its own (without automatic braking) reduces these crashes by 27%.
A recent international study found:
-
lane-keeping assist delivered the largest crash-rate reduction, especially for some of the most severe crashes (19.1% reduction)
-
automatic emergency braking significantly lowered rear-end and intersection crash rates (10.7% reduction)
-
blind spot monitoring cut lane-change and merging collisions moderately (3.5% reduction)
-
adaptive cruise control, surprisingly, appears linked to higher crash rates in some studies (8% increase), possibly due to drivers relying on it too much, using it in complex traffic, or paying less attention. Conventional cruise control was associated with an even higher increase (12%) in crash risk.
The most disliked feature is the most effective one
Lane-keeping assist is the most frequently disabled feature. Nearly 45% of drivers report they have turned it off – even though it is designed to prevent some of the most common and deadly crash types: run-off-road and lane-departure collisions.
Lane-keeping assist uses cameras to detect lane markings and gently apply steering inputs to keep the car centred. In practice, this can feel like the wheel is “tugging” or “nudging” against the driver’s hands.
On winding roads, in roadworks, or when markings are faded, the system may issue frequent warnings or attempt corrections that feel unnecessary. For some drivers, this sensation is disconcerting, and it’s a big reason why the feature is often switched off.
In most modern cars, lane-keeping assist can be switched off during a drive, but it will usually switch itself back on the next time the vehicle is started. Manufacturers design it this way to meet safety regulations and retain high crash-safety ratings from bodies like the European New Car Assessment Programme.
Why drivers switch them off
Despite clear safety benefits, many drivers reject advanced driver assistance systems.
Drivers cite a mix of frustrations and mistrust. False alarms, constant beeping, and perceived “over-correction” make some systems feel intrusive. Others switch them off after a single bad experience, or because they don’t understand the system’s limits – for example, that lane-keeping assist won’t function on poorly marked or unsealed roads.
Lane‑keeping assist is not always perfect, either. Testing by the Australasian New Car Assessment Program indicates some vehicles deliver abrupt or jerky steering inputs that are hard to override, making drivers feel out of control.
Further behavioural research into advanced driver-assistance system use has identified more issues.
Many drivers learn about these systems through trial and error and experimenting rather than formal instructions or manual.
What’s more, manuals are hard to follow. Many require above-average reading levels, use cluttered layouts, and lack standardisation across brands.
Many car owners also leave the dealership without knowing about their new car’s safety features. Many car sales staff have limited training on advanced driver-assistance systems and their limitations.
Further, the same feature can have a dozen or more different names across brands. Motorists acknowledge these systems aren’t always intuitive and show strong interest in learning how to use them safely.
Road design and traffic conditions also have a significant impact on trust and engagement. Less regular drivers avoid activating systems in complex or unfamiliar situations, while frequent drivers are more willing to experiment.
Some groups, such as young male drivers who see themselves as technically sophisticated, often have lower objective knowledge of these features but higher confidence in their knowledge.
What now?
Evidence from both international crash-data analyses and behavioural research points to several priorities for policy, industry, and driver education.
Safety regulations and purchasing incentives should prioritise the most effective features: lane-keeping assist, driver monitoring and autonomous emergency braking.
Some studies show neutral or negative safety impacts from adaptive and conventional cruise control. These features should be paired with targeted driver education and accurate marketing to set realistic expectations.
Consumer training should be improved at the point of sale. Standardised demonstrations and training can help bridge the knowledge gap.
Better documentation may also help. Simplified, standardised guides and in-car digital tutorials should be adopted.
The evidence is clear: advanced driver-assistance systems can prevent crashes, but the effectiveness of these features depends as much on human factors as on the technology itself.
*Milad Haghani, Associate Professor and Principal Fellow in Urban Risk and Resilience, The University of Melbourne; Akshay Vij, Associate Professor, UniSA Business, University of South Australia; Ali Ardeshiri, Senior Research Fellow in Urban Economics, University of South Australia, and Zahra Shahhoseini, Research Fellow in Public Health, Monash University
This article is republished from The Conversation under a Creative Commons license. Read the original article.
13 Comments
News flash....people dont like a lack of )personal) control...regardless of the consequences
I always keep that stuff on unless I am on country roads. The one that annoys is driver distraction. Traction control is awesome. When travelling I like to put on cruise control and stick behind someone else, then I can look around a bit while the vehicle will brake if the traffic slows.
PS when I say country roads I mean gravel where speed is limited to around 50 kmh due to single lane issues.
IMO, tech may be lulling drivers into a false sense of security, leading to complacency when it comes to important defensive driving skills. Also, technology can't eliminate sheer stupidity.
With technology, I like manual transmission. I also like cruise control on well engineered roads but never on the hilly, twisty, sometimes narrow, sealed and unsealed roads that are my more typical driving surfaces.
Active, defensive driving style is really important, as is reading road surface, an important skill I try to apply all the time to avoid extraneous matter, potholes, and the worst corrugations.
I don't like lane assist. From the start of my driving experience, keeping to the left was drummed into me. Many sealed roads have quite a wide strip of seal to the left of the side white line. On left hand corners, I use that as then it gives greater safety separation between my vehicle and oncoming vehicles.
But I guess there's a difference between the rural roads I mostly frequent and congested city streets and motorways. I always have to anticipate a 50+ tonne truck and trailer coming towards me, probably with the trailer crabbing over the centre line (painted or notional) of the road.
Lane-keeping assist is a pain because it gives too many false positives. Only once out of hundreds of times that LKA has beeped at me, has it been remotely close to (maybe) saving me from a collision.
Lane-keeping assist that keeps one in the centre of a lane, can be less safe than me moving left or right within a lane, to keep away from the vehicles in other lanes that aren't keeping within their lane/side of the road! But it still wants to beep at me for moving closer to a painted line, while it is oblivious to the fact that I'm doing so to move away from another drifting vehicle or, as LouB mentions, the possibility of an oncoming vehicle crossing the centre line on a corner.
The biggest issue with drive assistance technologies more broadly is that they can't predict or anticipate the way a human driver can and should. So, they are always late and reactive - a bit like an inexperienced, learner driver.
While driver assistance systems are a good idea in theory, and while single function things like ABS and seatbelts work well, other parts of their their implementations can leave a lot to be desired as they break the prime directives of good design.
Good human-machine systems are a conversation, where both the human and the machine can feed back and affect each other's behaviour and one control maps to one thing. Poor human-machine systems are a one-way lecture that expects humans to adapt to the system, no matter how opaque and plain dumb it might be, and there is a lot of nonsensical clutter around the essentials.
We resented that second approach in our gym teachers back in the day. Not that much has changed.
Don Norman, the grand old man of Human Centred Design, is at: https://jnd.org/ It might give you an even more jaundiced view of the current state of design - which much of it deserved.
I'm looking for a replacement car and after driving a few, the inanity of many of the systems is off-putting, annoying and occasionally scary.
Common behaviours that are symptomatic of the increasing intrusion of shouty, one-way systems in our vehicles -
- Panic automatic emergency braking in the middle of two lane roundabouts when someone, still safely in their lane, passes in front of you. That one nearly caused an accident on a couple of occasions.
- And the same where the AEB reacts to parked cars(!) on Dunedin's narrow, winding streets.
- Endless chirping and and wheel twitches from false positives by the lane keeping because our road marking is poor and our patchwork quilt road surfaces means the lane keep assists shy away from illusions like a skittish horse: at least you can teach a horse that things are OK.
- Driver attention systems that have a fit if you wear sunglasses, look in to a wing mirror, or scratch your nose.
- Satnav systems with ineptly designed, hard-to-use interfaces that are permanently out of date and are a chore to update. Why, in the age of Google maps, is proprietary satnav still so poor?
- Keyfobs and no mechanical key unless you pry off a cover plate to get at it. I surf and electronic fobs and salt water do not mix. Same would go with pretty much any outdoor activity in this country.
- Vast number of 'features' built in that simply add complexity for no discernible reason and hide the most commonly used things. Do we need nested menus to tune a radio station or see how far we've come today?
There are also the legal issues around the (mis-)use of the things your vehicle collects on you, like acceleration, braking, location and other data: https://www.cbsnews.com/news/gm-selling-driver-data-car-insurers-texas-… Can you see our data being protected by motor vehicle companies who see a saleable item?
Thought not.
And there's security. A lot of the driver-attention systems collect biometric data: they have cameras that watch you and some can recognise different drivers. Biometric data is problematic, as while you can change a password when (and it's when, not if) your data gets hacked, that's not an option for biometric identification. https://en.wikipedia.org/wiki/Automotive_hacking
Why are well-engineered, simple, efficient vehicles now apparently impossible to make? I think I get why the kids pick up the older Japanese cars and restomod them: they're comprehensible and fun - characteristics in short supply in modern cars.
"Restomod" is very good. Stolen.
Restomod is by no means original on my part. For a working definition...https://www.velocityrestorations.com/blog/what-is-a-restomod/. And before you ask, I'd love one of the Broncos https://www.velocityrestorations.com/for-sale/. Even if the price makes my eyes bleed.
I got a bit carried away with my comment, but it boils down to: when people actually perceive the value in the systems they'll use them.
I don't think being constantly, causelessly scolded is helping that perception.
"when people actually perceive the value in the systems they'll use them."
Over the years in farming there's always a multitude of people and organizations complaining how backward farmers are. In my experience farmers always apply your sentence and most often they've tried the new tech and found it wanting. If it works then it's snapped up pronto......and used for some completely different purpose to it's design ;)
I'm also currently considering upgrading my car and I'm really in two minds. Do I go for a 4-5 year old top of the line gas guzzling European, or for a brand new Korean or Chinese high-end electric or hybrid car ? The first seems to appeal to my heart, the second to my reason.
Why not a 4-5yr old Japanese? Mind, if you intend top sell on the euro in the next few years then sure, but post 120-150k European cars are still lacking in durability in comparison.
Euros can be fiendishly expensive to fix, and parts for anything slightly out of the ordinary, you'll have to wait for.
If you're not terribly hung up on driving involvement and just want something very comfortable and extremely reliable, buy a Lexus.
It has the bonus of avoiding the badge issues around some of the Euro and US brands as well, and car brand these days doesn't mean a lot. I'd cite Stellantis - https://www.stellantis.com/en/brands, while Volvo is owned by Tata the Indian Steel and car firm, and Rolls Royce by Volkswagen. Car brands are just corporate bargaining chips, but now mean little in terms of quality - https://www.warrantywise.co.uk/blog/news/toyota-triumphs-in-warrantywis… - or enjoyment.
Just drive everything and buy what meets your needs, and you like the feel of.
I rented a new car in Switzerland recently with plenty of driver aids. I hated it, it was like a nagging passenger. It would "bing" if I went 1 kph over the speed limit, physically turn the steering wheel if I was not in the middle of the lane and, worst of all, it had a camera on my eyes at all time and it would tell me to focus on the road if I didn't look straight ahead.
European standards can be useful in many ways, but clearly it is over-encroaching in this space. Too much of the state telling you what is best vs giving you the tools and leaving it up to you.
We welcome your comments below. If you are not already registered, please register to comment.
Remember we welcome robust, respectful and insightful debate. We don't welcome abusive or defamatory comments and will de-register those repeatedly making such comments. Our current comment policy is here.