No, not like that, that was last week. But if you’re old enough to remember Back to the Future the first time round, you probably remember Knight Rider, a TV show about David Hasselhoff fighting crime with a sentient talking car called KITT, in which the Hoff managed to be the less plausible human being of the two. Last week brought us a step closer to this dream becoming a reality when Tesla released a software update that introduced many of its customers to “Autopilot”, the first real elements of what will one day become driverless cars.
Over the past few years elements of so-called “driver-assist” technology have been introduced to us, things like parking sensors, cruise control and most recently auto-parking. Technically, this update is just an addition to that same driver-assist category. Tesla is keen to point out that these cars don’t drive themselves and that people should keep their hands on the wheel at all times, but it feels like so much of a bigger leap than that because, well, just look at it.
Perhaps realising that this is the future, most car makers, as well as outside players such as Google, are now developing some kind of automated vehicle and it is widely expected that fully driverless cars will be ready for general sale by the 2020s, but you might start to see them popping up on your street well before then. In July 2015 the British government set out its self-described “light-touch” code of practice for testing driverless cars which does not restrict where the vehicles can drive, nor require testers to obtain permits.
So what technology would a car need to make to jump to fully driverless? A GPS system, much like a modern satnav or phone uses would let the car know where it is and allow it to work out the best route to a given destination. Radar and sonar systems would be utilised by sending out radio or sound waves respectively and using the signals which bounce back off solid objects to work out how far away they are and how fast they’re moving. Radar works over longer ranges and is able to see through adverse weather conditions such as fog, snow or dust, whereas sonar has the advantage of being able to detect both hard objects like cars and walls as well as softer objects like animals and, it comes as a relief to say, pedestrians. Cameras can be deployed in a number of locations on the car to detect information not visible to radar or sonar. Low cameras keep track of the lane markings and help the car stay in lane. Higher front-facing cameras with image recognition can read road signs such as speed limits and stop signs. Cameras work less well in extreme weather such as snow. Finally, a number of on-board systems are needed to control the car’s response such as automatic braking, accelerating and steering. For a better look at how all the data these systems gather is turned into a driving experience, Google’s Chris Urmson gives a good demonstration of the calculations made by the car. I’ve started it at the relevant point but the whole video is worth watching.
It still sounds ridiculously futuristic to say that fully self-driving cars will be here in less than ten years, but once we reach that tipping point (and it’s getting close) all new cars will be like that. I guess the only thing I can really compare it to is the rise of the smartphone. Pre-2007 all that phones could do was make calls, send texts and maybe take a grainy, low-res dickpic that would cost a quid to send by MMS. In fact they were so bad for sending dickpics we hadn’t even invented the word dickpic yet. Then the iPhone came out and suddenly we’re here less than a decade later facetiming, translating in real time and swiping on Tinder whilst we sit on the toilet. And once Apple did it, everyone kept pace with them to stay relevant. Well, not quite everyone. RIP Nokia.
But that’s where we are now with cars, in the flip-phone era, and once self-driving cars arrive it’s not difficult to see the near future unfold in front of you. Driverless cars will almost certainly have some way of communicating with each other, allowing them to avoid roads which are becoming congested or, if they can’t avoid congestion, to maintain a slow and steady speed to keep traffic moving, rather than the irritating stop-start of today’s traffic jams.
The number of cars in existence would almost certainly drop. Living here in the past we largely need to own our own cars and keep them close by so we can get to them easily when we need to drive to work. But if the car could drive itself to your door and pick you up, why would you need to own one at all? Subscribing to some sort of driverless car-Uber app would surely become the norm and once you’ve been dropped off at work the car can run other people around town all day before picking you up again in the evening. Parking lots and multistories would become a thing of the past and the space could be used to build something worthwhile. (My suggestion would be affordable housing but I doubt anyone will listen.)
Safety would in all probablility improve as well. In 2014, 311 billion miles were driven on Britain’s roads, 79% of which were driven by cars and taxis and 1,775 people were killed as a result of road traffic accidents, 94% of which were caused by human error. Cars which spot and respond to danger quicker than humans will improve road safety. Google does occasionally suffer minor collisions during its ongoing testing phase and details them all in its newsletters, though almost all the accidents are caused by human drivers rear-ending Google’s cars.
Public transport would also likely take a hit. Whilst most intercity train routes would still be a quicker travel option, the ability to work, sleep or relax during transit means that driverless cars offer many of the benefits of train travel but with the added bonus of privacy and without needing to get to and from a station. Unless the railways are renationalised (and perhaps even if they are) the high price of train travel may also turn many off. Bus services may also see a decrease in use as the young, elderly and disabled who cannot drive could still use a driverless car. This could also be a great social leveller, allowing those sections of society who currently can’t get around enjoy the same independence and opportunities as the rest of us.
So where’s the catch? Whilst safety will be improved, Google’s experience shows that accidents may still occur. As Tesla was at pains to point out, for current driver-assist cars, liability remains with the driver, but whilst car manufacturers will be reluctant to get on board with it, they may find themselves liable once fully driverless models arrive. And yes, whilst accidents will happen it’s important to see the improvement on today’s situation rather than fixate on the outlying data point. As with all new upstart technologies that threaten the bank balances of the established order, a backlash will be quietly fomented and we’ll happily buy into it. Matthew Inman from The Oatmeal said it best when he wrote:
“Even if in a few years self-driving cars are proven to be ten times safer than human-operated cars, all it’s going to take is one tragic accident and the public is going to lose their minds. There will be outrage. There will be politicizing. There will be hashtags. It’s going to suck.”
In that moment it will be important to remember just how bad humans are at appreciating risk. We constantly fail to do it properly, which is why everyone is acutely aware of the few plane crashes that happen, despite the fact you’re more likely to die driving to the airport than on a plane.
A more serious threat might arise from cars being increasingly connected to the internet (remember Tesla’s latest update happened over Wi-Fi) making hacking a real concern, especially if the car is being driven by a computer. Some pretty strong safeguards are going to have to be designed into these vehicles and with companies like Google involved, a robust cyber-defence should be achievable. Let’s just hope no car manufacturers – like, oh I don’t know, Volkswagen maybe – manage to find a way to cheat their tests.
I for one can’t wait to use a driverless car and I’m fully on board with the idea that I might never own a car again. It won’t mean the end of driving “properly”, it will just move it on to a leisure activity. Going out for a spin on the country roads or enjoying a track day will still be possible in the world of future cars, but it will come without the price of having to negotiate the ring roads at rush hour. And when you bring Siri into the car with you to keep you entertained on your driverless commute, you might wonder if, in a small way at least, we haven’t all become David Hasselhoff.