Self-driving cars: Are they really the future?

Formerly seen as a just vision of a possible future – near or far – by our favourite science-fiction films, self-driving cars are becoming increasingly non-fiction.

While consumers themselves haven’t necessarily requested such an advance in automotive technology and mobility, it has been spurred on by governments, manufacturers and tech companies alike as some sort of natural progression for motoring.

Consumers’ willingness to use such vehicles might not necessarily be on the agenda (for now), but incredible amounts of money is being spent to make it a reality, and a safe one at that. 

Who is developing self-driving cars?

As mentioned above, this isn’t just about your standard car manufacturers developing such technology. In fact, if anyone is really pushing forward with self-driving cars, it’s tech giants such as Uber the Google-backed Waymo and Apple who are chomping at the bit to get these vehicles on the roads.

It should be no surprise, really, that companies who have vast expenditure on hiring drivers to run their network of vehicles are the ones primarily looking to have driverless cars. 

In regards to the manufacturers who are looking to join this endeavour, you’ll find the likes of BMW, Ford and Nissan among the ones most intent on adding such technology to their cars. With Uber, meanwhile, Volvo have become the US company’s automotive partner in their push for autonomous cars being added to their fleet.

Other manufacturers have taken a similar route as Volvo, coupling up with delivery and ride-sharing companies such as e-Palette (Toyota & Mazda). BMW, on the other hand, has joined forces with Intel and self-driving tech company Mobileye to ensure its autonomous cars are safe and intelligent.

Like Intel, fellow chip company Nvidia are also helping car makers such as Audi, Mercedes and Volkswagen join the race to have self-driving cars on the road.

Perhaps the most famous manufacturer who is currently on the road with at least some form of driverless technology is Tesla. On models like the Model S and Model X, Tesla’s Autopilot feature allows the car to drive itself on motorway-like roads, but the occupant is still required at the wheel at all times in case they are called into action.

How do driverless cars work?

To explain how self-driving cars operate, there are five core levels of autonomy set out by the Society of Automotive Engineers (SAE). Level 1 is seen as the most basic and Level 5 the most advanced.

- Level 1: this is a fairly straightforward affair and is a level of autonomy that has actually been around for quite some time now. Level 1 includes isolated features, such as radar cruise control and lane-keep assist; the former has been available since the 1990s. These features work through an array of cameras and sensors, but the driver is still there to drive the car.

- Level 2: this is pretty much where we stand today and includes the likes of advanced cruise control functions where the car can take over directional, throttle and brake capabilities. Thanks to the use of sat nav, this feature can read corners up ahead and brake as necessary on an automatic basis and also keep a set distance behind the car in front. This is also the level of self-driving cars that sees such abilities as automated-parking.

- Level 3: referred to as ‘conditional automation’ by the SAE, Level 3 is pretty much where Tesla is at with its Autopilot feature. This is next big stage of actual driverless cars, where a function can be applied by the physical driver for the car to enter a mode to have all aspects of driving done for you at certain points – but the driver must be at hand to take back control if required. Again, refined use of maps, radar and sensors are used in conjunction with environmental data to keep the car doing what it should be doing.

- Level 4: this is an area of driverless cars where the vehicles truly do drive themselves, but have their reach restricted through geofenced areas. These are also cars that you probably wouldn’t own, but merely ‘borrow’ to get around built-up areas and almost be like the shared pods seen testing in Milton Keynes over the last few years.

- Level 5: now to the point where a lot of companies are aiming for – full autonomous vehicles. Level 5 sees us take the previous step but derestrict its scope of destination, with full driverless capability brought about by hyper-fast algorithms crunching vast amounts of data at all times, making the cars fully sentient to what’s around them.

When will self-driving cars be available? Well, manufacturers are aiming for an evolution of Level 2 by the end of 2019, while the subsequent Levels will come not too long after that.

Level 4 cars are predicted to be available as soon as the early years of the next decade, with Level 5 arriving just a two-to-three years after that. By 2035, analysts (HIS) are forecasting circa 21 million driverless cars on the roads on a global scale.

What obstacles are in the way?

As touched on previously, one potentially obvious obstacle to driverless cars is: will anyone buy them?

Given the stat above of having millions of such cars on our roads, you think that isn’t a problem for those developing such methods of transport. However, with safety being a key issue here, how many people around you can honestly say they trust these vehicles?

For Uber, its preparations were halted in March 2018 after a Volvo XC90 it was testing in Arizona killed a pedestrian; however, it was seen to be predominantly the error of the driver who had turned off Volvo’s autonomous safety systems and was watching TV at the time of the incident. Despite this, there were concerns that Uber’s own system – not related to the Volvo system – did not make any apparent effort to slow the vehicle down or take evasive action at any point.

Meanwhile, Tesla has come up with its own issues in regards to how drivers are using its Autopilot technology. Multiple deaths have unfortunately occurred due to drivers not taking notice of warnings for them to take back control of the vehicle whilst in Autopilot, leaving questions as to whether we as drivers can be trusted with this technology, while the feature has also come under its own scrutiny.

Elsewhere, there are other problems that perhaps aren’t so obvious. On such factor is that in a place like the UK, mobile coverage can be hit-and-miss in places, especially in the countryside. This, as this article from The Guardian suggests, could be an issue for driverless car technology which depends so much on fast and reliable mobile data.

One of the biggest advantages of self-driving cars is said to be that our roads will be safer, despite the early troubles seen in testing, as the technology should be able to react to dangers and potential incidents quicker than a human. Assuming that’s the outcome, what happens to car insurance when these vehicles become mainstream?

While not a complete obstacle to driverless cars, it’s certainly a big question that needs answering in earnest – how do you insure a driver who doesn’t drive the car? With little to no data to work off, it will be hard for underwriters initially to set premiums accurately (if at all), but fortunately, there are groups already on the case.

An insurance product in the US, for example, is able to monitor drivers’ use of autonomous features and even help cut premiums through sensible use of the technology. However, if self-driving cars take off as some foresee, if nobody is driving, will car insurance need to evolve beyond the driver?

Self-driving car ethics

Now, this is an interesting, not commonly thought of proposition: can a driverless car can be ethical?

As this article delves into, in the unlikely circumstance of a driverless car needing to avoid an incident, what manoeuvres should be programmed into the software from a moral standpoint? Touched upon in the same article, the ‘Trolley Problem’ could stand out as a relevant example to driverless cars if they were to have make a decision based on an upcoming collision or incident.

While most will just hope a self-driving car will in general keep the occupants and those outside of the vehicle safe, what if there was a decision to made where the car had to choose between hitting numerous pedestrians or sacrifice the single occupant of the vehicle to save the many? 

Perhaps more pertinent would be, would a motorist be willing to enter one such vehicle knowing that decision might have to be made on their behalf? 

To avoid these types of dilemmas, you might say the person at the wheel could take control back from the car, but, as driverless cars get more widespread, might their occupants become less aware of their environment? It’s certainly one to think about.

Are self-driving cars the future?

It’s fairly evident that the answer to this question is a confident ‘yes’. Manufacturers and tech companies alike are in full motion to get their driverless cars on the road for both consumer and industry use.

However, with an estimated 21 million of these vehicles projected to be on our roads by 2035, that is drop in the ocean when considering the bigger picture. There are an estimated 1.4 billion cars on the road, so to have just 1.4 per cent of those be self-driving barely makes a dent into the industry’s infrastructure as a whole.

Despite this, it will be enough that changes will have to be made, whether that be legislation, infrastructure or approach to driving. Drivers of conventional cars will be aware of these sentient vehicles around them, while services such as taxis and car sharing will inevitably take advantage of this new set-up.

What will be interesting to see for many is if this move to driverless cars does indeed save lives in the long-run and if there will be any dangers in our own influence on these vehicles when ‘at the wheel’.