For all the expectation, excitement and energy surrounding them, autonomous cars still aren’t here. Sure, they’re being tested in almost every city from Mountain View to Bangalore, yet at the risk of pointing out the obvious, the prospect of 100% self-driving cars being sold to the public is still a long, long way off. There are semi-autonomous cars on the market, admittedly, such as the Tesla Model S, yet even if these could ever be fully trusted to drive entirely on their own, they face the seemingly intransigent obstacle of severe legal restrictions on cars that require no human input.
Yet this isn’t the least of it, for in the world of automobiles there’s another quickly emerging current which suggests that, in actual fact, the future of motoring isn’t going to be autonomous cars. This is connected-car technology, a rapidly developing set of devices and systems that enable vehicle-to-infrastructure (V2I), vehicle-to-vehicle (V2V) and vehicle-to-pedestrian (V2P) communication. Not only does such tech provide the motorist with real-time information that makes driving safer, but it also shows that his future resides less with self-driving BMWs and Fords, and more with connected cars and the intelligent road networks they communicate with.
The most recent example of V2I tech comes from Las Vegas. Here, motorists who purchase the 2017 Audi A4 will now find that the car essentially talks with the city’s traffic lights, which provide the vehicle with real-time info enabling its dashboard to display a timer for when the lights will turn from red to green. This might not sound too Earth-shattering, but in receiving live data from the Regional Transportation Commission of Southern Nevada (who oversee Las Vegas’ 1,300 traffic signals), Audi motorists are spared the impatience and rashness that can come from not knowing when lights will flip. As such, they’re empowered to drive on the road in a safer and more informed way.
This example of Traffic Light Information (TLI) shows the basic principle of how connected road networks — a combination of big data and smart cars — can improve a driver’s knowledge of the road and, by extension, her safety. Yet added to this, if we look at other areas in which connected-car tech is now appearing, it becomes apparent that it avoids or solves many of the problems that stand in the way of fully autonomous cars.
And it’s precisely because it avoids them that it — and the ‘intelligent’ drivers it enables — are much more likely to be the future of motoring than self-driving cars.
As mentioned above, perhaps the biggest obstacle facing autonomous cars is the law. In September 2016, for example, the United States Department of Transportation published a stringent 15-point checklist of safety conditions semi- and fully autonomous cars will have to meet before the Federal Government will allow them on the road. These include requiring companies to share the copious amounts of driving data they collect with the National Highway Traffic Safety Administration. They also require that all autonomous vehicles become subject to “pre-market approval” from the Government, and that the technical specifications of the software used in self-driving cars are also shared in advance of going live. Ostensibly, this is so the NHTSA can check to make sure that autonomous vehicles are safe, yet it’s a requirement that many jealous car makers and tech firms have resisted up until now, since much of their driving and software info is a proprietorial, closely guarded secret.
It’s therefore arguable that the necessity of sharing data will act as a block on innovation in the autonomous car realm, as intimated by David Strickland, the general counsel for the Self-Driving Coalition For Safer Streets, who in response to the new Federal guidelines affirmed the importance of “protecting the data rights and frankly the property of all the innovators and manufacturers.” Yet this problem with info sharing isn’t something faced by cars capable of V2I communication, since these are simply the same kind of driver-operated cars we’ve known for decades, from cars that use GPS systems to those that use radios for traffic updates. Such cars don’t require their manufacturers to open their data to governments and the public. Because of this, vehicle-to-x technology can be invested in knowing full well that any investment will see the maximum possible return and the maximum possible commercial advantage.
But there’s another legal hurdle confronting autonomous cars, a hurdle which isn’t so much a legal obstacle in itself as much as something which undermines the very rationale behind the consumer adoption of self-driving vehicles. As written in the Federal 15-point checklist, and as written into state codes from California to Nevada, drivers either have to be constantly behind the wheel of a self-driving car, or they at least need to be ready to take over in case something unexpected happens (as in Florida). This is a sensible obligation, yet at the same time it defeats the purpose of having an ‘autonomous’ car, demanding that drivers babysit their machines rather than sit back and take a nap.
Except in rare circumstances where driverless autonomous cars are allowed in ‘controlled areas,’ the law still obliges drivers to remain alert at all times, and if they have to remain alert, than they might as well be driving. Once again, this isn’t a problem affecting cars which simply communicate with road networks, since these were meant to be driven by a human in the first place.
More importantly, connected cars get the direction of babysitting correct, in that rather than requiring humans to watch over A.I. and data, it requires A.I. and data to watch over humans. This is the preferable order of things, for the whole point of introducing technology into motoring is to compensate for shortcomings in how people drive. If a human driver is required to take over from her autonomous car when something goes wrong, then she still ultimately remains the fundamental ‘guarantor’ or ‘underwriter’ of the vehicle, even though the whole idea of autonomous vehicles is predicated on the assumption that she can’t really be trusted to have such a role. Conversely, V2I cars assume that she can, and that all she needs are regular info updates and alerts from her surrounding environment to ensure that she drives as safely as possible.
And make no mistake, V2X tech can make motoring much safer. At Cadillac, for example, they’re working on vehicle-to-vehicle technology that will transmit a car’s position to other nearby cars, thereby providing them with early warnings on sudden turns or stops. Using such tech, cars will be able to receive moving images from nearby vehicles and display them on their digital dashboards, with the result being that human drivers are given the kind of fore-knowledge and intelligence that would otherwise be reserved for AIs. With this kind of intelligence, they’ll be more aware of their surroundings and the likely implications of their actions, and they’ll be correspondingly less likely to cause accidents.
In other words, connected-car tech will help them to solve the problem of motoring safety, without having to put them in an autonomous car that —because of legal and commercial concerns — is likely to be delayed for years to come anyway. This is why intelligent road networks are more likely to be the future of driving that autonomous cars, and even though v2I and V2V systems are in the early stages, there’s every sign that their development will overtake that of fully autonomous vehicles. For one, there’s the 5G Automotive Association that was formed by major manufacturers in September to facilitate the development of V2X technologies, and there’s also the Department of Transportation recent proposal to make V2V communications mandatory for all “new light vehicles.”
But there’s also the simple fact that connected-car tech is already being developed in parallel with research into autonomous vehicles, as can be seen with how Amazon has just patented a V2I system that will inform vehicles of when reversible traffic has, well, reversed. This suggests that, when the incipient self-driving-car industry hits the inevitable wall, the V2I, V2V and v2P technology being built around it will slip away and be used to power ‘normal’ cars like the Audi A4 or latest Cadillac. Increasingly, we’ll find vehicles that warn their drivers of possible collisions arising from lane merges or emergency stops, that indicate an approach to such riskier areas as construction sites, and that even signal to pedestrians (via smartphone) that a car is veering riskily close to them.
The upshot of all this will be to reduce the 30,000+ people who die on American roads each year, and all without requiring such a massive paradigm and legal shift as the widespread adoption of autonomous cars by consumers. With the (politically mandated) integration of connected-car technology into automobiles, they’ll be little need to overcome the resistance of tech companies to sharing their data, they’ll be little need to overcome the resistance of governments to permitting driverless cars to be sold to the general public, and they’ll be little need to overcome the resistance of drivers and drivers groups to giving up an enjoyable, self-affirming activity.
Yet the creation of intelligent road networks — comprising vehicles, infrastructure and pedestrians — will nonetheless make motoring a much safer pursuit. This is why we’ll be seeing many, many more cars like the latest Audi A4 long before autonomous vehicles are available to consumers, and it’s why motoring will continue reserving a space for human autonomy for a long time to come.