The vision of a self-driving car is one first rooted in science fiction; a notion that’s persisted in one form or another for decades. Our fascination with handing control of many aspects of our life to machines has never really waned, but we continue to focus on the easiest concepts to understand.
For example, autonomous vehicles are frequently talked about as an everyday reality of our future. With that, they’ll play an increasingly impactful role in society - changing everything from employment to urban design. A connected city filled with autonomous vehicles is one that requires significant infrastructural upgrades from those that we inhabit today.
With those changes will come massively improved safety on the roads (90% of accidents are caused by human error) and the accompanying fall in road traffic deaths - a figure pegged at 1.25 million by the World Health Organization in 2013.
Today, in the US, vehicle accidents are the leading cause of death of teenagers — in 2015, six teenagers died in car accidents every single day, according to the CDC.
Yet, to harness the potential direct safety benefit of self-driving cars, we need a lot more education on how to integrate them on the roads while there are still also human drivers behind the wheel.
Analysis of the data garnered in self-driving car trials suggests that self-driving cars are already safer than human drivers on average (not just teens) in at-fault accident terms, but were involved in more minor collisions. This is likely due to the small but subtle differences following the exact rules of the road.
As well as being safer, self-driving cars in cities of the future will reduce the need for car ownership.
Fewer cars on the roads means less congestion, and a reduced need for short and long-term car parking - if it’s your car, it could simply drive itself to a nearby location until you’re ready to leave. Spaces no longer need to be right where you’re headed, because machine-to-machine communication and beacons on parking spaces will make for more efficient usage.
In more optimistic visions of the future, this newly reclaimed space in cities will be transformed into pedestrian utopias and green spaces, but this feels like exactly that - an optimistic view.
While efficiency doesn’t always equate to a better standard of living, in this case, it does pose the ability to significantly reduce deaths - not just those directly related to driving, but also through things like improved air quality, for example. More broadly, self-driving - by their nature, electric - cars could have many more effects on pollution beyond solely reducing reliance on fossil fuels.
Safety we can’t yet see
More excitingly, perhaps, are all the changes that are harder to predict but potentially more revolutionary.
Cars themselves, for example, could undergo a fundamental shift in design as a direct result of a fall in collisions and injuries. Safety features that are currently mandatory for our error-prone human driving will be superfluous in a future where cars don’t crash. Naturally, this has knock on effects on the materials we use to build cars (and the parts that we no longer need) and a ripple effect of industries too.
The plethora of sensors on self-driving cars makes them prime candidates to be repurposed for other uses too - and not just as mobile weather station data collection agents. Some companies, such as Veniam, already putting plans in place to potentially use them as roving collection agents that could help detect and prevent crime by using on-board cameras. Naturally, the very suggestion of roaming crime prevention cameras won’t sit comfortably with people concerned about encroachment on privacy, but these are the sort of challenges that technological developments lead to in society, and answers that will need to be found.
The integration of data-driven driving into day-to-day life doesn’t has hundreds of potential benefits for our safety in a whole range of contexts. Imagine another example, where you’re being driven home through the dark of night in a torrential downpour. Not only does it no longer matters that human visibility is reduced to just a few yards, but you’ll also be automatically re- routed around any flooded roads thanks to data uploaded in real-time from sensors.
All of these potential effects on our safety and health are realistic expectations of when self- driving cars are commonplace, but none of them will be achieved without continual improvement to less headline-worthy technological developments.
In order to function, self-driving cars need a huge amount of real-time data, much of which will be drawn from the cloud and combined with artificial intelligence. Our parking and rain examples above rely on machine-to-machine communication and beacon infrastructure - all of which depends on solid connectivity in one form or another.
Where a person needs road signs, traffic lights and all manner of other ‘street furniture’ to get from A-B, the combination of Deep Learning and machine-to-machine communication means that one day, it’s likely none of this will be needed at all, but only if we continue to solve the less glamorous tough infrastructural and security problems that will underpin the driverless car revolution along the way.