The future is here: nuTonomy, a US-based start-up, has launched its first driverless taxi trial in Singapore after a test earlier in the year with empty cars.
Of course, being a trial, there is a driver present to take over, should any mishaps occur, but otherwise the cars are fully in control of themselves.
Exciting for those who like all things futuristic, but many are still baffled by the notion of a ‘self-driving’ car.
Are they really driverless? How safe is it? And how long will it be before small talk with taxis drivers is a thing of the past in the UK?
To find out, we sat down with Tyron Louw, a Research Assistant at the University of Leeds’ Institute for Transport Studies.
It is the first test of its kind
“As far as I know it’s the first trial of a self-driving taxi, but importantly it’s the first to tackle automated driving in a dense inner city,” says Louw.
We have all seen footage of driverless cars meandering slowly around simple test tracks, which proves a point but doesn’t really compare to driving on roads with a near infinite amount of subtle differences.
The race is on
Given the level of coverage that tech giants like Google and Uber, and car-makers like BMW, Mercedes and Ford have been receiving in their attempts to bring autonomous vehicles to the masses, you would think nuTonomy’s first steps into the driverless car race would be seen as a major PR fail for them.
But Louw argues that it is not a case of who is first, but rather who is best.
“It’s not about who gets it first, it’s about who gets it right. These companies are sitting on much more advanced automated driving systems than those being demonstrated by nuTonomy.
“Unlike nuTonomy, companies like Tesla have the customer base, manufacturing power, and capital to actually take them to market. nuTonomy has a lot less to lose if something goes wrong with the trial.”
It makes total sense that Singapore gets automated taxis first
Why has Singapore hosted the latest advance in a global technology race?
The circumstances around testing these kinds of automative advancements are a lot more political than you might think.
“All major projects requiring some investment in infrastructure need political will,” says Louw.
“In the UK, while we only have one major vehicle manufacturer in Jaguar Land Rover, the government has promoted the UK as an ideal location to demonstrate and test all manner of driverless technologies.”
Countries all over the world want the publicity boost of being the first place for driverless technology, and Singapore is no exception.
“Singapore sees the benefit in being associated with such advanced technology and forward-thinking,” says Louw.
We need to get our terminology right
Although the trial in Singapore deals in cars that can take care of themselves (albeit with a driver on standby in case of any accidents), that is because it is much easier to programme one city’s road network into a car’s system than it is the entire world’s.
Truly driverless cars that can operate literally anywhere could still be a long way off, and Louw suggest companies need to be clearer about their technologies.
“The confusion is why we are beginning to see accidents while drivers are using Tesla’s ‘autopilot’ feature, where drivers misinterpret how intelligent their car actually is. Names are important, and in our field we tend not to use the term ‘driverless’ because it causes such confusion.
“To be considered truly ‘driverless’, it needs to be able to take you from A to B without you ever having to intervene. If the driver is still required to monitor the road then it is only a semi-automated vehicle.”
We may have automated features in UK taxis soon
While truly driverless cars might be a way off, we could see something similar – where a driver or operator sits in on a machine that is more or less automated – debut in the UK in the next few years.
Louw thinks that Uber’s planned trialling of automated taxis in Pittsburgh in the near future is good news for UK users who may want a futuristic pick-up from the train station.
“I don’t think it would be too long before we have some taxis in the UK with autonomous capabilities, perhaps by the end of the decade.
“However, I can’t see fully autonomous taxis, where there is no driver, for at least a few decades. Driving is a hugely complex task; it’s not trivial that humans are able to do it, nor is it to make a machine do it.”
And it will be safe – for the most part
It’s easy to feel unnerved by the idea of putting your morning commute in the hands of your car, and with the amount of variables on the roads, it might be hard for the average person to imagine any sort of driverless car technology as being 100% safe.
It is a point that Louw agrees with too, but he argues that drivers need education on what their cars can and cannot do before sitting behind the wheel.
“What is most important for safety is that drivers understand the limits of their automated driving system. Does it recognise cyclists? Can it navigate roundabouts? It’s vital that drivers know what they are responsible for.
“A vehicle programmed to work on a fairly standard motorway should do fine, but take it to a rural road then it might struggle to find the lane markings it uses to track the road.
“AI will eventually make these systems much more reliable, but for the moment we can only rely on these vehicles to do what they say they can do.”
Tyron Louw is a Research Assistant and PhD Student at the University of Leeds’ Institute for Transport Studies.