Driving Science · Car technology · Guide 44
The skills your car quietly deletes
A pupil passes the test in an instructor's 2023 car. It chimes when the wheel drifts over a white line, it puts a camera view on the screen the moment reverse is selected, and it lights an amber triangle in the door mirror whenever something sits in the blind spot. Six weeks later that same driver buys a 2014 hatchback for four and a half grand, and it has none of it. The question here is not whether the technology is good. It is whether the habits those sensors were quietly performing on the driver's behalf ever got built, and what happens on the day the sensors go away.
Section 1
The handover nobody plans for
Irish learners practise in two cars: an instructor's, usually newish, and a parent's, usually not. Then they buy a third. Nobody audits which safety systems each of the three has, and that is where the risk collects.
SIMI puts the average age of the Irish car fleet at 9.9 years across roughly 2.47 million cars, petrol at 11.3 years and diesel at 10.3. In early 2008 the fleet averaged 5.8 years. So the realistic first car for a newly qualified driver here was built around 2016, and it predates every mandate in this guide.
The gap is widening. Ireland registered 85,203 new cars in the first half of 2026, up 4.2%, while used imports ran to 47,109, up 39%, with June imports alone up 42.7%. Imports are usually older and may be built to another market's specification, so the fastest-growing part of the second-hand market is the part where you can assume nothing about what is fitted.
That last figure deserves a plain statement rather than a spin. 85 people died on Irish roads in the first six months of 2026 against 77 in the same period of 2025, and the RSA told the committee that over 600 people have been seriously injured so far this year. Nearly half were vulnerable road users and almost a third were 25 or younger. In 2025, 190 died, the highest since 2014. The cars are getting cleverer and the Irish outcome has not followed. Read no causation into that in either direction, but it is the context.
Section 2
What actually landed on 7 July, precisely
You will have seen headlines about a camera in your car watching your face. The honest version is narrower and more boring, and the details are what decide whether the concern is reasonable.
From 7 July 2026, every newly registered car, van, lorry and bus in the EU, meaning category M and N vehicles, must have an Advanced Driver Distraction Warning function. It has applied to new vehicle types since 7 July 2024. The legal basis is Regulation (EU) 2019/2144, the General Safety Regulation, supplemented by Commission Delegated Regulation (EU) 2023/2590, adopted on 13 July 2023. Motorcycles are not in scope. Neither are small-series manufacturers building in limited numbers, nor certain special-purpose vehicles: the UK government's reading of the regulation lists motorcaravans, ambulances, hearses and wheelchair accessible vehicles as outside it. For an ordinary mass-produced car or van, though, it applies. As EU law it applies in Ireland automatically.
The function activates above 20 km/h and must work in daylight and darkness. It maps where the driver is looking into zones, and the one that matters is Area 3: anything below a downward plane 30 degrees under the driver's horizontal eye line, which is where you would be looking at something distracting in your lap, such as a handheld phone. The UK Department for Transport's explanatory memorandum on the regulation gives the threshold in one sentence: "Warnings will be triggered if the driver's gaze exceeds 3.5 seconds in that area when travelling above 50km/h, or after a period of more than 6 seconds if travelling slower." The warning must be visual, plus either haptic or acoustic.
🔬 Three corrections to what you have probably read
It warns, it does not brake. Nothing in this function takes control of the car, applies the brakes or steers.
It is not facial recognition and it does not send video anywhere. The function must work without biometric identification, including facial recognition, and it operates as a closed loop: the data it needs is processed and retained on the device in the vehicle. The regulation does not require footage of the driver to be recorded or transmitted to a manufacturer, an insurer or a police force.
It does not apply to your existing car. There is no retrofit obligation. If your car was registered before 7 July 2026 it is unaffected, and at a 9.9-year average fleet age the great majority of Irish cars will not have the function for a decade or more.
One more distinction, because it is the commonest error in the July 2026 coverage. Driver Drowsiness and Attention Warning is a separate, earlier requirement that applied to newly registered cars and vans in the EU from July 2024, and it is about drowsiness. Advanced Driver Distraction Warning is about visual distraction, and it is the one that arrived this month.
The law describes a function, not a camera
The regulation specifies a gaze-monitoring function and the test procedures it must pass. It does not name a component. In practice nearly every manufacturer meets it with an infrared camera on the steering column, the dash top or near the interior mirror, because that is what tracks where a driver's eyes point. So the accurate sentence is that the common implementation is a camera, not that the law mandates one. Several outlets have got that backwards. The cameras were arriving before the deadline anyway, for commercial reasons: the same memorandum notes that "driver monitoring systems contribute to a vehicle's EuroNCAP score and this will add a further incentive to fit them to vehicles made for the GB market", a market that has no such requirement at all, and Euro NCAP's 2026 protocol requires continuous eye and head tracking, linked to assistance sensitivity, for its highest ratings.
The link to a law you already know. Area 3 is, in plain terms, the phone-in-the-lap gaze pattern. Using a handheld phone while driving in Ireland is a fixed charge of 120 euro plus 3 penalty points. The practical effect of the new function is that the car now chimes at exactly the thing the Gardaí already fine you for. The offence is set out in full in our Irish mobile phone driving law guide.
Section 3
The system you already resent, and why the complaint is half right
Distraction warning is arriving into a driver population that has already spent four years being nagged about the speed limit. To predict how people react to the new camera, look at what they did with the last mandate.
Intelligent Speed Assistance became mandatory on new car models sold in the EU from July 2022 and on all new cars sold from 7 July 2024, in an overridable form. That same July 2024 date brought a bundle of other requirements for all new cars: reversing cameras or sensors, drowsiness attention warning, emergency stop signal, lane keeping systems, automated braking systems and event data recorders. Note what is not on the list. Blind spot detection is required for buses and trucks, not for cars, so a blind spot monitor in a passenger car remains a manufacturer's choice.
Most people assume the speed system does more than the rules require. It must comprise a speed limit information function, plus either a warning function or a speed control function. Permitted feedback includes visual plus cascaded acoustic warning, visual plus cascaded haptic warning, or accelerator pedal force feedback. As The Car Expert puts it, "Most ISA systems do not physically prevent the car from going faster", and the pedal-pushback versions give way if you press harder. Every mandated version must be overridable and must let you switch it off for the current journey, and it normally comes back on when you restart the engine.
Why it gets Irish limits wrong
The system reads the limit from a forward-facing camera and from stored map data. Both fail predictably: dirty or obscured signs, out-of-date maps, roadworks, temporary limits, and signs meant for an adjacent road that the camera reads anyway. Ireland then added a national-scale version of the map problem. On 7 February 2025, under the Road Traffic Act 2024, the default limit on rural local roads dropped from 80 km/h to 60 km/h. That covers a very large share of Irish road length, so a car working from stale maps, or an import whose data was never built for Irish local roads, can be confidently wrong by 20 km/h over and over.
The complaint is legitimate on accuracy. Where it stops being legitimate is the conclusion. The car being wrong about the limit does not make you right. You are responsible for the sign, and the fixed charge and the points land on your licence, not the manufacturer's. The rules, including the February 2025 changes, are in our complete guide to Irish speed limits. There is a middle path most drivers skip: find the mute or override control before you need it, so you are not scrolling a touchscreen menu at 80 km/h, then turn the volume down rather than the function off. A system that is switched off is protecting nobody, and as Section 5 shows, that is not rhetoric.
⚠️ These switch-off figures are British, not Irish
Which? surveyed 1,584 UK car owners in January 2025 and published in March 2025. The percentages who turn features off at least some of the time: speed assist 46%, lane keep assist 42%, automatic emergency braking 34%, attention or drowsiness monitoring 32%, blind spot monitoring 30%. Overall, just over half of the owners who had one of these features turned at least one of them off at least some of the time. Separately, IAM RoadSmart's 2024 Road Safety Report, based on 2,008 drivers, found 34% never use adaptive cruise control despite having it, 25% use it often and 16% rarely.
Both samples are British. We could not find an equivalent Irish survey, so read these as an indication of behaviour in a comparable market, not as an Irish measurement. Note too that Which?'s own wording is "at least some of the time", which is softer than saying half of drivers routinely disable speed assist. The tidier 54% that circulates in trade coverage comes from the Driving Instructors Association's write-up of the survey, and it says "regularly disable". Which? itself publishes "just over half", so we have used that.
Section 4
What sixty drivers said they had stopped doing
This is the core of the guide, and it has to be presented as what it is: interview evidence about what drivers report, not a measurement of anybody's observation skill.
Soyeon Kim and Oscar Oviedo-Trespalacios published "Behavioural adaptation in Advanced Driver Assistance Systems (ADAS) use" in Applied Ergonomics volume 136 (2026), article 104799. It is a qualitative analysis of interviews with sixty drivers in Australia, and four interrelated themes came out of it: degradation of driving skills, shifting risk perception and tolerance, reduced cognitive and physical engagement, and diverse responses to system warnings. The authors report that over-reliance and complacency are common and can undermine the safety benefit the systems were fitted to deliver. Participants described declines in attention, decision-making and driving skill as they leaned on the system, and the warnings theme covered both dependence on warnings and warning fatigue.
The examples that stick with instructors are the ones to label most carefully. Drivers described no longer turning their heads to check a blind spot once a light in the mirror was doing it for them, and struggling to park without a camera view. Those are things participants said about their own habits. They are not before-and-after measurements, there is no control group, and self-report about your own competence is among the least reliable measurements in human factors. Take them as hypotheses worth teaching around.
The companion paper matters as much and gets less attention. Kim and Oviedo-Trespalacios, "Disuse of advanced driver assistance systems (ADAS)", Journal of Safety Research volume 95 (2025), pages 180 to 188, interviewed drivers of SAE Level 2 vehicles and found nine themes across three domains: the driving task, the human and the environment. Drivers cited "discomfort with system aggressiveness, lack of trust in detection capabilities, and incompatibility with their driving styles".
Switching a system off is usually not stupidity. It is a rational response to a device behaving in a way the driver cannot predict. The fix is a better system and a driver who understands it, not a lecture about safety.Reading Kim & Oviedo-Trespalacios (2025) on why drivers disengage assistance systems
Which? collected the same sentiment more viscerally. One driver in Greater Manchester described a lane keeping intervention in her Hyundai i20 like this: "To feel the steering wheel suddenly spin and the car begin to swerve was terrifying." That is one driver and one car, and it is not a claim about lane keeping assist in general. But it explains the switch-off statistic better than any appeal to driver responsibility does. Somebody frightened once by an intervention they did not understand will disable it, and telling them the fleet data is good will not change that.
▲ The claim
"All this technology is making people worse drivers. My nephew can't reverse without the screen and never looks over his shoulder. They should teach it properly instead of putting sensors on everything."
▼ The challenge
Two halves need separating. The skill-fade half rests on interview evidence, not measurement: sixty Australian drivers reported this pattern, consistently and plausibly, but nobody has demonstrated a decline in their actual observation performance. The "technology is bad" half runs into the crash data in Section 5, which is genuinely strong and points the other way. Forward collision warning with automatic emergency braking cut front-to-rear striking crash rates by 50% in US fleet data. Those are real crashes that did not happen.
● The verdict
Both are true at once, which is the useful conclusion. The systems reduce crashes at fleet level, and drivers report letting go of habits the systems appear to cover. The habits still matter, because the driver will change car, borrow one, hire one, or sit a driving test in which direct observation is the assessed standard. Train the habit and keep the system on. Choosing between them is the mistake.
Section 5
Where the technology plainly works, and the finding that decides how much
The fleet-level evidence has to be stated fairly, because it is good. Nothing in this guide is an argument for buying a car without these systems.
The cleanest numbers come from US insurance-claim and police-report data analysed by Jessica Cicchino for the Insurance Institute for Highway Safety, published in Accident Analysis & Prevention in 2017. Forward collision warning on its own cut rear-end striking crash rates by 27%. Low-speed automatic emergency braking cut them by 43%. The two together cut them by 50%. For rear-end crashes involving injuries the reductions were larger: 20%, 45% and 56%. The lane and blind spot systems are more modest but real. Blind spot detection lowered all lane-change crash rates by 14% and injury lane-change crashes by 23%; lane departure warning cut the relevant single-vehicle, sideswipe and head-on crashes by 11% and injury crashes by 21%.
The European Commission's projection for the whole General Safety Regulation package is that the mandated systems are "expected to help save over 25,000 lives and avoid at least 140,000 serious injuries by 2038". Read that for what it is: a forward projection for a bundle of measures, not a measured outcome for any single system.
Now the sharpest finding in the literature, and the reason Section 3 laboured the mute button. IIHS attributes lane departure warning's smaller measured benefit partly to drivers switching it off, while drivers with blind spot monitoring overwhelmingly leave theirs on. That is direct evidence that the switch-off rate changes a system's measured effectiveness. It is not a side issue or a matter of taste. It is the variable that decides how much of the benefit reaches the road.
🧪 The optimistic counterweight
A purely gloomy story about drivers rejecting safety technology would be wrong. IIHS found that by 2024, 87% of vehicles observed had lane departure prevention switched on, against 51% having any lane maintenance system activated eight years earlier. The lead author, Aimee Cox, put it plainly: "The results reflect a combination of better designs and a growing acceptance of crash avoidance systems, more generally". Systems that behave predictably get left on. That is the design lesson and the teaching lesson in one.
Complacency is not laziness
Naming the mechanism stops this becoming a moral argument about other people's character. Raja Parasuraman and Dietrich Manzey reviewed it in "Complacency and Bias in Human Use of Automation: An Attentional Integration", Human Factors 52(3), 2010, pages 381 to 410. Three findings are directly relevant. Complacency occurs under multiple-task load, when manual tasks compete for attention. Attention paid to the automated task is driven in part by trust in the automation. And it appears in both naive and expert participants, and is not overcome by simple practice. That last part is why "concentrate harder" is not a plan. What helps is giving the habit a fixed trigger instead of relying on remembering it, which is what Section 9 is about.
This site has covered the risk side already: our article on risk compensation and the Peltzman effect has a section on the driver-assist trap. Keep the three ideas distinct. Risk compensation is a change in a driver's appetite for risk, complacency is the attentional route by which a safety gain gets spent, and skill fade is a change in the driver's repertoire. This guide is about the third.
Section 6
What none of this proves
The claim here is modest on purpose. If the evidence supported a stronger one we would make it, so the small print goes in full rather than in a footnote.
🧪 The honest small print
Nobody has shown that assistance systems measurably degrade driving skill. The core paper is a qualitative interview study of sixty self-selected drivers in one country. It records what those drivers say about their own habits. It does not measure observation performance, it has no control group, and sixty interviewees are not a population. Simulator research on manual skill after periods of automated driving is genuinely mixed, with some studies finding no significant after-effects on driving performance. Treat skill fade as a well-argued hypothesis with plausible mechanism support, not a demonstrated effect.
There is no crash-reduction evidence for the new distraction warning function. None. Every effect size quoted here belongs to automatic emergency braking, forward collision warning, blind spot detection or lane departure warning. The government memorandum on the distraction regulation says so itself: "its overall impact is difficult to establish given the subjective nature of driver distraction and its relationship with road traffic collisions."
Even the size of the distraction problem is a wide range. The European Commission's road safety knowledge base states that distraction plays a role in 5 to 25% of crashes in Europe and calls that likely an under-representation. An in-depth study cited there, Thomas et al. 2014, found distraction a contributory factor in 8.3% of fatal crashes and inattention in 5%. We have deliberately not used the 10 to 30% figure that circulated in July 2026 press coverage, because we could not trace it to a primary Commission document.
Some real-world associations run the wrong way. An inventory of real-world assistance-system performance by Haghani, Vij and colleagues reports lane keeping assist at −19.1%, automatic emergency braking at −10.7% and blind spot monitoring at −3.5% on relevant crash rates, but adaptive cruise control at +8% and conventional cruise control at +12%. We could only read those figures through an explainer written by the authors, the underlying paper was not accessible to us, and the inventory draws on literature of mixed quality. These are associations. It is not evidence that adaptive cruise control causes crashes, and we are not claiming it does.
And the house position on training. This site already publishes that formal driver training has not been shown to cut crash risk, with the Cochrane review landing at a rate ratio of about 0.98: see does driver training actually work. Nobody here is going to tell you a lesson about car technology will save your life. What a lesson can do is make you competent in a car that has no sensors, and make you a driver who knows what their own car will and will not do.
Section 7
What the Irish driving test actually says about car technology
Almost nobody writes about this part, and it comes straight from the RSA's own documents rather than from opinion.
The RSA's Making Your Mark: Marking Guidelines for the Driving Test contains two sentences that settle most of what learners ask. On parking aids it is blunt: "The use of 'Park Assist' is not permitted." On cameras it is permissive but conditional: "Use of video or audio devices whilst reversing is acceptable, if used in addition to the required observation."
Read the second twice, because the operative words are in addition to. The camera is allowed as an extra and never as a substitute. The same guidelines define direct observations as observations to the front, the sides and the rear, including blind spots, which is exactly the habit a blind spot monitor is most likely to erode. So the assessed standard in the Irish test is the habit the technology most readily replaces. A tester is watching your head and your eyes, not your screen.
⚠️ The internal-camera question is genuinely unresolved
The RSA's Standard Procedures for Conducting Driving Tests (2025) states at page 13: "Where a vehicle has factory fitted internal cameras, the Tester must be satisfied that these cameras have been deactivated for the duration of the test." The tester may ask the candidate to demonstrate that they are deactivated.
Now hold that against the new distraction warning function. Whether the driver can deactivate the warning, or the whole system, is left to the manufacturer. Some cars will let you. Some may not. We found nothing published that resolves how the deactivation requirement interacts with a distraction warning system the manufacturer does not let the driver disable. So this is a question, not a rule. If you are presenting a 2026 car for test, ask your instructor in advance, know what internal cameras your car has and what can be switched off, and expect the tester to raise it. Do not let anyone tell you there is a settled rule here, because there is not one yet.
One thing the same document does settle: vehicles "presenting with cameras in place of external mirrors" are acceptable, provided they work effectively and show sufficient observation to the required areas.
Both RSA documents are quoted verbatim above from the official PDFs. The rsa.ie site returns an HTTP 403 to automated fetchers, so the links in the sources section point at the right documents but will not open for a bot. For the wider set of questions learners ask on the morning of a test, our test day questions guide works through twelve of them from official RSA procedures.
Section 8
Buying your first car in Ireland: what you will and will not get
Model ranges and individual cars are two different things, and on the second-hand market only the second one matters.
Given the fleet numbers in Section 1, the realistic first car has no speed assistance, no distraction warning, no blind spot monitor, and a reversing camera only if a previous owner paid for the option. The table below is the mandate timeline for cars, from the European Commission's list of required systems and the distraction regulation dates. Treat it as a floor, not a spec sheet: manufacturers fitted many of these earlier and voluntarily to chase Euro NCAP points, and equally an individual car may have had an option box left unticked.
| Registration era | What EU rules required on a new car |
|---|---|
| Before July 2022 most of the Irish fleet, including the average 2016 car | None of the systems in this guide were required. Anything fitted was a manufacturer decision or a paid option. Check the individual car, never the model range. |
| New vehicle types from July 2022, all new cars from 7 July 2024 | Intelligent Speed Assistance; reversing camera or sensors; driver drowsiness attention warning; emergency stop signal; lane keeping system; automated braking system; event data recorder. |
| New vehicle types from 7 July 2024, all new registrations from 7 July 2026 | Advanced Driver Distraction Warning, the gaze-monitoring function described in Section 2. |
| Never required on cars at all | Blind spot detection, mandated for buses and trucks only. Also adaptive cruise control and any form of park assist. If a car has these, somebody chose to fit them. |
Four checks before you hand over money
- Ask what this specific car has, then verify it. Sit in it, work through the settings menus and find the systems by name. A model range that offered blind spot monitoring tells you nothing about the car in front of you.
- Check the warning lights for those systems are out. A permanently lit lane keeping or camera fault light means a system fitted and not working, which is the worst of both worlds: you will assume you have it and you will not.
- Ask whether the windscreen has been replaced, and whether the forward camera was recalibrated afterwards. Speed assistance and automatic emergency braking usually look through that glass. We could not find a verified Irish price for recalibration, so take no figure from anyone without a quote, but do ask whether it was done.
- Read a Euro NCAP assisted driving grading for what it measures. It grades assistance competence and safety backup, and rewards a system that keeps the driver engaged. It is not a cleverness score, and a higher grading does not mean a car that needs less from you.
Two notes on imports, since they are the fastest-growing part of the market. Great Britain has not adopted the distraction warning requirement, though manufacturers generally build one specification for both markets, so a GB import may or may not have it: check the individual car rather than generalising. And an older import may carry map data that has never heard of Ireland's 2025 rural speed limit change.
Section 9
Six habits to train anyway
None of these is a philosophical position about technology. They are the observation and control habits the Irish driving test already assesses, so training them is simply what passes.
- The shoulder check, on a trigger rather than a reminder. Decide now what makes you turn your head: mirror, signal, shoulder, move, every time. The fixed trigger matters because Parasuraman and Manzey's work says you cannot rely on remembering under load. A light in the door mirror is a second opinion, not a first one.
- Mirrors on a rhythm, not on a prompt. Work them on your own cycle, before you need the information. A driver who looks only when prompted has outsourced the timing of their own observation.
- Know where the back of your car is without a screen. Learn reference points through the mirrors: where the kerb appears, where the number plate behind you disappears. Then use the camera to confirm what you already decided, which is the order the RSA guidelines require anyway.
- Know 60 km/h by feel. Engine note, wind noise, how fast the hedge goes past. If your only route to your own speed is a number on a dial, you have nothing when the dial is wrong. Our guide to what the car is telling you covers reading the vehicle instead of the display.
- Read the road for the limit, not the dash. The sign is the law. The car's guess is a guess, and on Irish rural local roads since February 2025 it is often wrong by 20 km/h.
- One deliberate no-camera manoeuvre per lesson. Reverse, park or turn about with the geometry decided by your eyes and mirrors. Cheapest insurance there is against the day you drive a car with nothing.
For more on the observation side, building observation skills is the longer treatment and inattentional blindness explains why looking and seeing are not the same thing. The twelve mandatory lessons are covered in EDT lessons explained.
A real teaching problem worth naming. A learner permit holder may not drive unaccompanied and may not use motorways, so almost all supervised practice happens in the instructor's car or a family car. Those two often have completely different levels of assistance, and the car the learner buys is usually a third one, older again. Then N plates go on for two years and the supervision stops. That sequence argues for training the manual habit in whichever car has the fewest sensors, not the most.
Section 10
For instructors: when to introduce the technology
There is now published research on this exact question rather than trade opinion, and it is worth knowing what it does and does not say.
Soyeon Kim, Simeon Calvert and Marjan Hagenzieker published "Exploring ADAS driver training in driving academies: Perspectives from driving instructors" in Safety Science volume 194 (2026), article 107042, interviewing fourteen instructors across four European countries who had delivered structured training on assistance systems.
Four findings transfer. Training built confidence and discouraged over-reliance, so teaching the systems is not the same as encouraging dependence on them. The instructors valued combining explanation, controlled practice off the public road and real-road driving. They judged that ADAS training should come after a novice has mastered fundamental driving skills, not alongside them: no interval appears in the published abstract, so the honest version of that finding is a sequence, not a timetable. And they called for standardised ADAS training, with the leasing companies, dealerships and regulators pulled in alongside the driving schools, which is a systemic point rather than a classroom one. They also flagged that older drivers need a tailored approach.
🔬 Four practical notes for an Irish ADI
Teach each system by name and by limitation. "It reads white lines, and it has nothing when the lines are worn or the road is wet and shining" is more useful than a demonstration of how clever it is.
Demonstrate a false intervention where it is safe to do so. A pupil who has felt a nuisance intervention with you beside them, on a road you chose, will not be frightened into disabling the system alone six months later. Which?'s interview material is a catalogue of drivers who met their first intervention on their own.
Show pupils the controls before they need them. Mute and override, in their own car, parked. Fighting a touchscreen at speed is a distraction you created.
Keep the manual habit as the assessed standard, and cover the test mechanics. The marking guidelines require direct observation to the front, sides and rear including blind spots. Park Assist is not permitted in the test. A reversing camera is permitted only in addition to observation. A pupil presenting a new car should know what factory-fitted internal cameras it has and what can be deactivated. Materials sit in the ADI Hub.
Section 11
What to do this week
Learning to drive: name every assistance system in whatever car you drive, out loud, with your instructor. Decide your shoulder-check trigger and use it for the next four lessons until it stops being a decision. Do one deliberate reverse per lesson with the screen ignored.
Buying a new car: before you drive it off the forecourt, find the speed assistance and distraction warning controls, and find out whether the distraction warning can be switched off in this particular car, because that is the manufacturer's choice and not the law's. Then read the two pages of the handbook that cover the assistance systems.
Buying a first car for your teenager: ask what safety systems this specific car has rather than what the model offered, and check no warning lights for them are lit. Then absorb the arithmetic: at a 9.9-year average fleet age, the car you buy will be roughly a decade behind the car your child learned in. Our guide for parents and the N plate rules cover the rest of that first solo year.
Instructing: take two things from Section 10. Demonstrating a nuisance intervention somewhere controlled, and showing pupils the override controls, give the best return for the time they cost.
If any of that is easier with somebody in the passenger seat, that is what we are for. Kris and the team cover south-west Dublin, Kildare and Meath, with instructors in Tallaght, Navan and D15. Call 087 394 8102.
The verdict
Keep the systems on. The fleet evidence for automatic emergency braking, forward collision warning and blind spot detection is among the better evidence in road safety, and the clearest single finding in it is that the benefit shrinks when drivers switch things off. A muted system is protecting you. A disabled one is not.
Then train the habits as though the car has nothing, because sooner or later it will. The skill-fade evidence is honestly weaker than the headline suggests, and there is still no published crash evidence at all for the distraction warning function that arrived this month. But you do not need proof of skill fade to justify a shoulder check. The Irish test already requires direct observation to the front, sides and rear including blind spots, and the car you own in five years will very probably be older than the one you learned in.
The uncomfortable part of the Irish numbers is how slowly all of this arrives. A 9.9-year fleet age means these mandates take a decade or more to reach the average Irish car, and 85 people died on Irish roads in the first half of 2026 alone. The technology in new cars is not what will decide your first year of driving. What you do with your eyes is.
Sources & further reading
References
- Commission Delegated Regulation (EU) 2023/2590, adopted 13 July 2023, supplementing Regulation (EU) 2019/2144 as regards Advanced Driver Distraction Warning. The correct primary citation for the technical requirements. Note: EUR-Lex returned empty content to our fetcher on repeated attempts by both the HTML and PDF routes, so we cite it rather than claiming to have read it. eur-lex.europa.eu
- UK Department for Transport, Explanatory Memorandum on Commission Delegated Regulation C(2023) 4523 (ADDW). Source for the mechanics quoted here: approval required for all new types from 7 July 2024 and all vehicles from 7 July 2026 (paras 7 to 8); the Area 3 definition and the "gaze exceeds 3.5 seconds ... above 50km/h, or after a period of more than 6 seconds if travelling slower" threshold, plus the visual and haptic or acoustic warning requirement (para 5); "its overall impact is difficult to establish given the subjective nature of driver distraction and its relationship with road traffic collisions" (para 12); and "driver monitoring systems contribute to a vehicle's EuroNCAP score and this will add a further incentive to fit them" (para 14, where the sentence is written about vehicles made for the GB market). Also the scope: all mass-produced M and N category vehicles (para 4), with small-series manufacturers and vehicles such as motorcaravans, ambulances, hearses and wheelchair accessible vehicles outside the requirement (para 16). An authoritative government reading of the regulation, but a secondary reading of it. assets.publishing.service.gov.uk (PDF)
- European Commission, DG GROW, 5 July 2024. The systems mandatory on all new motor vehicles sold in the EU from 7 July 2024, having applied to new vehicle types from July 2022: intelligent speed assistance, reversing cameras or sensors, driver drowsiness attention warning, emergency stop signal, lane keeping systems, automated braking systems and event data recorders for cars and vans; blind spot detection technologies required for buses and trucks. Also the projection: "They are expected to help save over 25,000 lives and avoid at least 140,000 serious injuries by 2038." single-market-economy.ec.europa.eu
- Seeing Machines, Understanding Advanced Driver Distraction Warning (ADDW) systems. Supplier explainer on Regulation (EU) 2023/2590: activation above 20 km/h, operation in daylight and darkness, function without biometric information including facial recognition, closed-loop on-device data handling, and the point that manufacturers choose "whether it is possible for the driver to manually deactivate the ADDW warning and/or the ADDW system". Vendor source, corroborated where load-bearing. seeingmachines.com
- Commercial Vehicle Workshop News (Ireland), 21 July 2026. Irish trade coverage of the requirement taking effect, and the distinction between Driver Drowsiness and Attention Warning (required from July 2024) and Advanced Driver Distraction Warning. cvworkshop.ie
- ETSC, Intelligent Speed Assistance. The EU agreed in 2019 to make an overridable version of ISA mandatory on new models of car sold in the EU from July 2022 and on all new cars sold from July 2024. ETSC's position is that ISA should be on by default and that any switched-off state should last only until the vehicle is restarted. etsc.eu
- The Car Expert, What is intelligent speed assist and can you turn it off? Plain-English account of the speed limit information function plus warning or speed control function structure, the permitted feedback types, why camera and map-based limit detection goes wrong, and the switch-off-per-journey behaviour. "Most ISA systems do not physically prevent the car from going faster." Consumer motoring site, not a regulator. thecarexpert.co.uk
- Which?, Is technology ruining new cars? Survey of 1,584 UK car owners, January 2025, published March 2025. Turn off at least some of the time: speed assist 46%, lane keep assist 42%, automatic emergency braking 34%, attention or drowsiness monitoring 32%, blind spot monitoring 30%. Which?'s own summary of the overall figure is "just over half of drivers who have an ADAS feature on their car reporting they turn at least one feature off at least some of the time", which is the wording used in this guide. Source of the Greater Manchester i20 driver's account of a lane keeping intervention. The rounder 54% figure, and the stronger word "regularly", come from the Driving Instructors Association's write-up rather than from Which?. UK data, not Irish. which.co.uk · driving.org
- Fleet News, on IAM RoadSmart's 2024 Road Safety Report (survey of 2,008 drivers). 34% never use adaptive cruise control despite having it, 25% use it often, 16% rarely. We did not open the IAM report itself. fleetnews.co.uk
- Cicchino, J. B. (2017), Accident Analysis & Prevention, via the IIHS bibliography. Forward collision warning alone cut rear-end striking crash rates 27%, low-speed automatic emergency braking 43%, and the two together 50%; for rear-end crashes with injuries, 20%, 45% and 56%. iihs.org
- IIHS, Lane departure warning, blind spot detection help drivers avoid trouble (Cicchino). Blind spot detection lowered all lane-change crash rates 14% and injury lane-change crashes 23%; lane departure warning cut relevant single-vehicle, sideswipe and head-on crashes 11% and injury crashes 21%. Also the finding that lane departure warning's smaller measured benefit is partly attributable to drivers switching it off: the page states that US drivers frequently turn lane departure warning off, and estimates the benefit would match earlier studies had it been left on. The companion point that blind spot monitoring is overwhelmingly left on appears in secondary coverage of the same IIHS research rather than on this page, so treat it as the weaker half of the pair. iihs.org
- IIHS, Fewer drivers are opting out of lane departure prevention (2024, lead author Aimee Cox). 87% of vehicles observed had lane departure prevention switched on, against 51% with any lane maintenance system activated eight years earlier. "The results reflect a combination of better designs and a growing acceptance of crash avoidance systems, more generally". iihs.org
- Kim, S. & Oviedo-Trespalacios, O. (2026). "Behavioural adaptation in Advanced Driver Assistance Systems (ADAS) use." Applied Ergonomics 136:104799. Qualitative analysis of interviews with sixty drivers in Australia. Four interrelated themes: degradation of driving skills, shifting risk perception and tolerance, reduced cognitive and physical engagement, and diverse responses to system warnings. Note: ScienceDirect returned HTTP 403 to our fetcher, so the design and themes here come from the published abstract as indexed. sciencedirect.com
- Kim, S. & Oviedo-Trespalacios, O. (2025). "Disuse of advanced driver assistance systems (ADAS)." Journal of Safety Research 95:180-188. Interviews with SAE Level 2 drivers; nine themes across three domains (driving task, human, environment). "Drivers cited discomfort with system aggressiveness, lack of trust in detection capabilities, and incompatibility with their driving styles." research.tudelft.nl
- Kim, S., Calvert, S. C. & Hagenzieker, M. (2026). "Exploring ADAS driver training in driving academies: Perspectives from driving instructors." Safety Science 194:107042. Fourteen instructors across four European countries. Training builds confidence and discourages over-reliance; value in combining theory, controlled practice and real-road driving; training should come after novices have acquired fundamental skills; older drivers need a tailored approach; and a call for standardised ADAS training with cross-sector collaboration among leasing companies, car dealerships and regulatory bodies. No timing figure appears in the abstract, so none is stated in this guide. research.tudelft.nl
- Parasuraman, R. & Manzey, D. H. (2010). "Complacency and Bias in Human Use of Automation: An Attentional Integration." Human Factors 52(3):381-410. Complacency occurs under multiple-task load when manual tasks compete for attention; attention to the automated task is driven in part by trust; it appears in both naive and expert participants and is not overcome by simple practice. Cited at abstract level. journals.sagepub.com
- RSA, Making Your Mark: Marking Guidelines for the Driving Test. "The use of 'Park Assist' is not permitted" (p.21) and "Use of video or audio devices whilst reversing is acceptable, if used in addition to the required observation" (p.20); direct observations defined as observations to the front, sides and rear including blind spots. The correct primary citation, quoted verbatim from the official PDF. Note: rsa.ie returns HTTP 403 to automated fetchers. rsa.ie (PDF)
- RSA, Standard Procedures for Conducting Driving Tests (2025). "Where a vehicle has factory fitted internal cameras, the Tester must be satisfied that these cameras have been deactivated for the duration of the test" (p.13), with the tester able to ask the applicant to demonstrate deactivation; vehicles "presenting with cameras in place of external mirrors" acceptable where they work effectively and show sufficient observation (p.16). Quoted verbatim from the official PDF; rsa.ie is also 403 to fetchers. rsa.ie (PDF)
- ERSO / European Commission, Thematic Report: Main factors causing fatal crashes. "It is generally estimated that distraction plays a role in 5 - 25% of crashes in Europe", described as likely an under-representation; Thomas et al. (2014) found distraction a contributory factor in 8.3% of fatal crashes and inattention in 5%. Crash-risk multipliers: about 1.5x for hands-free phone use, about 2.5x for operating vehicle systems, 3 to 3.5x for handheld phone use and over 6x for texting. Also: drivers' compensatory behaviours such as slowing down and increasing headway are "mostly ineffective in counterbalancing the impaired reaction time". road-safety.transport.ec.europa.eu (PDF)
- SIMI, Ireland's National Vehicle Fleet. Average national car fleet age 9.9 years across roughly 2.47 million cars, up from 5.8 years at the start of 2008; petrol 11.3 years, diesel 10.3 years. simi.ie · SIMI new car registrations release, June and July 2026: 85,203 new cars registered in the first half of 2026 (up 4.2%) and 47,109 used imports (up 39%), with June 2026 imports up 42.7% year on year. simi.ie news
- The Journal, 15 July 2026, reporting RSA evidence to the Oireachtas transport committee. 85 people killed on Irish roads in the first six months of 2026 against 77 in the same period of 2025, with over 600 seriously injured; nearly half vulnerable road users and almost a third aged 25 or younger; 190 deaths in 2025, the highest since 2014. Used because rsa.ie is not fetchable. thejournal.ie
- RTÉ News, 5 February 2025. The default speed limit on rural local roads reduced from 80 km/h to 60 km/h on Friday 7 February 2025 under the Road Traffic Act 2024. rte.ie
- Euro NCAP, Assisted Driving gradings and 2026 protocol changes. Assistance competence and safety backup as the graded dimensions, with the framing that the driver remains responsible and engaged; the 2026 rating overhaul into Safe Driving, Crash Avoidance, Crash Protection and Post-Crash Safety, with the highest ratings requiring continuous eye and head tracking linked to assistance sensitivity. euroncap.com · euroncap.com press
- Haghani, Vij, Ardeshiri & Shahhoseini, in The Conversation, 26 August 2025. Inventory of real-world assistance-system safety performance: lane keeping assist −19.1%, automatic emergency braking −10.7%, blind spot monitoring −3.5%, adaptive cruise control +8%, conventional cruise control +12% on relevant crash rates. The explainer names driver monitoring as one of the systems worth prioritising but gives no percentage for it, so none is quoted here. The underlying ScienceDirect paper returned 403, so we could not read its methods; the inventory draws on literature of mixed quality and these are associations, not demonstrated causal effects. theconversation.com
Related on this site: Risk compensation and the Peltzman effect · Does driver training actually work · Building observation skills · What the car is telling you · Used car buying checklist · What new drivers aren't taught · Driving Science hub