“I think it will be useless, out of date in 10 or 15 years’ time. This is something that is not going to be required, it shouldn’t be planned.” Those are the words of Dermot Desmond when discussing the oft-planned, oft-delayed Dublin Metro, the underground, overground (wombling free?) rail service which has been rumbling through minds, if not actually the ground, for a quarter of a century now.
The fact that Dublin is a major, heavily populated capital city without a proper urban rail service remains something of a national disgrace, and a major drag on both the quality of life of those of us who live in the city and environs, and doubtless something of a boat-anchor when it comes to investment. That is not in question.
However, Mr Desmond’s comments that such a service would be superseded by autonomous cars? That’s another story, and it makes one think of the rush to rip up Ireland’s rail network back in the 1950s and 1960s, because apparently cars and buses were the way to go. What we would do to have those rail lines magically back again now.
The question, though, is can autonomous cars truly replace and supplant a proper public transport infrastructure? The answer is a very messy “probably not”.
READ MORE
Autonomous technology has been the darling of tech investors for a couple of decades now, ever since the US Defence Advanced Research Projects Agency (DARPA) first started asking universities and car makers to create robotic cars that could drive themselves. DARPA’s idea was to create military trucks that wouldn’t need human drivers, thereby putting fewer people in harm’s way and allowing the deployment of troops to more pressing tasks in time of war.
The concept didn’t take long to percolate through into a potential business opportunity for those looking to provide a service to regular commuters. Why not relieve people of the mundanity of traffic jams? Why not improve on-road safety because computers – in theory – make fewer mistakes than unpredictable humans? It sounds tempting.
However, the end product is still somewhat underbaked. The likes of Google’s Waymo, and Lyft, and Uber have all put robotic cars on to roads, offering them as taxis for the tech-obsessed, but while the systems have been shown to work, they have also been shown to be easily confused, and prone to accidents, especially when confronted with the requirement to share road space with human drivers.
If you could wipe all other cars off the road, and replace them with fleets of robotic taxis, would that make an adequate replacement for a Metro service? Possibly, but then the evidence of a robotic-car traffic jam in San Francisco seems to show that even computers get confused at times, and no system is without glitches. Equally, we’ve already had our first computer-car-caused death, when a cyclist in Arizona was mowed down by an experimental Uber robot car.
That leads to questions about how autonomous software would cope with being asked to share streets with existing human drivers, as it would have to at least in the interim. One industry expert suggested to The Irish Times that any robotic car unleashed on to the streets of Beijing would soon be going nowhere as canny local drivers would work out that they could barge in front of it, forcing its sensors into an endless loop of yielding.
There’s also a potential environmental hiccup. Prof Hannah Daly of the school of engineering and architecture in UCC told Business Ireland magazine: “Like any new technology, whether autonomous vehicles ultimately lead to better outcomes depends on how they’re used and who controls them.
“If they’re integrated into the transport system in ways that make transport more accessible, efficient and environmentally friendly, they could have real benefits. But if they simply encourage more residential sprawl, reduce investment in public transport or remain a luxury for the well-off, then they risk locking us into even more car dependency and entrench an unequal and environmentally damaging future.”
And that’s the problem with autonomous cars – they’re largely being marketed as trinket travel. Look at Tesla’s Robotaxi – an autonomous car with only two seats (one of which is, for the moment, occupied by a human supervisor who’s there to mash the emergency stop button if needs be) and with no access for disabled users is no sensible replacement for a proper public transit system – it’s a rich person’s toy and little more.
Renault is one of the first car makers to break ranks with the autonomous hegemony. While the French car-making giant is still working on robotic driving tech, it’s doing so to develop it for use in public transit systems.
In fact, Renault – working with autonomous transport start-up WeRide – operated a robot bus for the French Open Tennis championships this year, shuttling fans to the tennis stadium at Roland Garros, or back to the car park via the Place de la Porte d’Auteuil. The bus itself was based on a Renault Master van chassis.
It is a bus, though, and this is Renault’s reckoning – that autonomous tech makes great sense for big public transport schemes, where the vehicles have dedicated lanes and set schedules, but far less sense for private cars, mostly because the French car maker reckons that the cost of such systems will always be too high for most punters to stomach.
“In the case of individual vehicles, Renault Group is concentrating its efforts on the Level 2 or even Level 2+ level, with several driving assistances that are at the top level of the market and make its vehicles safe and pleasant to drive with confidence, such as contextual cruise control or lane-keeping assist, or soon the automatic overtaking function. Although assisted, the driver remains responsible for driving” says Christophe Lavauzelle, Renault’s corporate communication manager.
“There is a significant technological complexity gap between level L2 automation and level L3 autonomy, because the vehicle must be able to operate safely in complex environments with limited driver supervision. At this stage, the induced cost to be borne by customers, in relation to the driving benefits, would make demand insufficient or even anecdotal.
“On the other hand, when it comes to public transportation, Renault Group sees the relevance of offering autonomous vehicles, with an annual need estimated at several thousand minibuses over the next few years.”
At the same time, the group is making sure that the architecture of its vehicles can evolve towards the autonomous car if expectations, regulations or the cost of technologies make this breakthrough feasible.
“Renault Group is moving forward to implement its autonomous vehicle strategy. As a result, thanks to our experiments and our partners, the best in their fields, we will be in a position, well before the end of this decade, to propose a highly relevant range of autonomous, low-carbon minibuses to meet the growing needs of the regions” says Gilles Le Borgne, Renault Group’s chief technical officer.
The rest of the industry seems to be in a mood to acknowledge harsh realities as well, according to a report from S&P Global: “The hype about fully autonomous cars dominating our streets has faded. This shift was evident at the 2025 Consumer Electronics Show (CES) in Las Vegas, once a key platform for autonomous vehicle companies to shine.
“This year’s announcements, however, reflect a growing industry trend: rather than setting lofty goals for fully self-driving cars on any road, tech companies and traditional automakers alike are now prioritising practical, market-ready solutions. As expectations between these two groups converge, the future of autonomous vehicles will be defined by incremental progress and collaboration.”
So, the answer is no, autonomous cars powered by AI are no substitute for a proper, public, mass-transit system.
