The 7 myths of automated driving
Much is said of what automated driving is and what it will change. Less is said about what it is not and what it will not change. Here, your fears of a future with robots ruling the roads are put to rest.
A comment from a reader of an automated driving article says: “OVER MY DEAD BODY will I let a robot drive my ‘67 Chevy Camaro.” Admirable devotion to their most prized possession indeed – but perhaps a little far-fetched. Or is it?
Indeed, our vehicles are evolving to unprecedented levels of intelligence and self-reliance; to a point where “the system” will be able to take care of driving as you sit back and relax. So, it is not the concept of non-human “drivers” that is far-fetched – but rather the image that this comment conjures up: a collection of nuts and bolts and circuitry in a quasi-human form saying “Must. Drive. Car.” as it clambers into the front seat. Or the fact that existing cars will suddenly be made autonomous – by virtue of some retrofitted “system” being enforced.
Although not all misconceptions about automated driving are of this magnitude, myths about the technology certainly abound. Here’s the top seven – and an attempt to put them right.
1: “You can buy a fully automated car today”
Dreamy commercials and YouTube videos may lead you to believe otherwise but as of 2016, there is no real fully automated car on the market. A potentially fatal misunderstanding, as the Tesla crash in May 2016 reminded us: a beta-stage Level 2 vehicle was misunderstood to be a full-blown Level 3 car.
However, you can buy a car that can exhibit semi-automation in certain instances. Take parking for example: more and more premium cars are able to park themselves – with no driver in the driving seat. However, the driver is always watching close by from the outside. Still though, none of today’s production vehicles recognize all road signs and signals, communicate directly with other vehicles or allow you to cruise hands-free while engaged in another task.
That is not to say such technology isn’t in development. The road to Level 5 – fully automated driving - currently has two lanes. One takes the evolution approach where more and more advanced driver assistance systems are added to and developed until the car becomes fully automated. The other takes the revolution approach – heading directly to an autonomous vehicle.
Steering away from passenger cars for a moment, institutions may invest in other types of automated vehicles that are already available for certain protected environments like a university campus.
2. “Automated driving means the end of driving for pleasure”
If your idea of a perfect Sunday is getting into your sports car and hitting those long and winding country roads for a few hours then don’t worry. Your keys will not be confiscated anytime soon! It is safe to say that the focus for the development of automated driving is certainly not one of denying people the pleasure of driving.
Indeed, the vast majority of car manufacturers have even vowed to keep this passion alive by saying that the driver will have a choice of driving modes – automated for the tedious stuff and manual for the fun! So the real challenge is to develop driverless cars that are able to co-exist with manually driven cars in our traffic systems.
3. “Autonomous cars will be programmed to decide over life and death”
Of course, the idea of programming a machine to make life or death decisions raises difficult ethical questions. Take the “trolley problem” scenario: a conundrum where a run-away train is heading for five people who are stuck on the track. The controller of the train cannot brake but can divert the train onto another track where there is only one person stuck. It is a lose-lose situation that can easily be extended to a self-driving car. What if the car must go one of two ways and there are human lives at stake? What option does it take? Should it, for instance, choose to harm two elderly people over two toddlers?
An ethical crossroads indeed. However, experts see a way out: the goal is to develop autonomous cars that are able to prevent such rare scenarios from ever playing out. The car of the future must be connected to all the infrastructure, vehicles and people around in such a way that it will foresee potential dangers far ahead of time and thus avoid ethical dilemmas from happening. A challenge for sure – but not an insoluble one.
4. “Automated driving removes all human error from driving”
What a world it would be if this were the case! Alas…unfortunately it won’t be. As long as there is human input in a process, there will always be room for human error.
What automated driving will do is gradually take over more of the driving task to reduce this room. Even if vehicles become fully autonomous, they will need to be programmed – by humans! So no; automated driving is not the cure for us making mistakes. However, it will go a long way towards eliminating the horrific consequences of making them while driving.
5. “Statistics prove it: Humans are bad drivers”
Road traffic accident statistics are worrying: 90% of all fatal car accidents are caused by humans. Even one death on our roads is a tragedy. But relatively speaking, their incidence is actually extremely low (e.g. in 2015 in the U.S. there were 7.1 fatalities per 1 billion km driven). A testament to how good we really are as drivers, as experts confirm. Good indeed – but there’s room for improvement. And that’s where automated driving comes in.
The technology will complement the impressive human skill set by eradicating the incidents that are the result of our weaknesses (for instance, a driverless car will never get distracted by a smartphone). And it will do so while reducing emissions and making driving a lot more comfortable. So don’t feel too put out: right now, driverless cars are still learning how to master complex driving tasks from humans. And it will take a long time until it’s the other way round.
6. "Automated Driving is being developed mainly for the consumers market”
Today, you may be led to believe that automated driving is all about individual cars for the end consumer. While private passenger cars have been stealing most of the headlines, in reality, commercial applications are much more likely to be the key drivers of the revolution.
Take truck platooning for example – where a number of automated trucks equipped with driving support systems closely follow one another. The trucks are driven in convoy by smart technology. This is set to have a huge impact on the transport industry in the future in terms of safety and cost. Some 4,000 people die each year in the US alone from trucking accidents: platooning could reduce that significantly since the room for driver fatigue and error is lessened. Not to mention cost: companies can cash in on savings from not having drivers in all trucks as well as reduced fuel consumption.
Another application is driverless public transport systems for urban spaces – like this impressive vision for Singapore. And lest we forget, Uber have already started testing autonomous cars in their endeavor to meet CEO Travis Kalanick’s vision of a driverless fleet by 2030. Why? Because no drivers means freeing up 75% of the revenue!
7. “If I’m reading this, I won’t live to see fully automated driving on the road”
This one is a little harder to debunk. Why? Because it depends what age you are when reading this! But really, it depends on semantics. What do people mean when they talk about fully automated driving?
These days, they’re mostly referring to Level 4 automation – which we are on the brink of achieving. This would see production cars that are able to drive autonomously on highways, for example. But once you get off the highway, you’d have to take over manual control again. To those who think bigger, those who imagine a driverless world, in which you could even summon your autonomous car to pick you up after a big night out, we’d have to say: stay tuned! That scenario requires reaching Level 5. Driverless cars would have to be able to master all traffic scenarios without a human present. Achieving that goes well beyond technical challenges and requires multiple other challenges to be resolved.
Putting a timeline on this becomes somewhat less of an exact science! Right now, we can say that Volvo, for instance, is unleashing a fleet of XC90s which will drive autonomously in certain use cases in 2017. BMW claims that it wants to bring the fully automated BMWi Next into series production as early as 2021. So let’s hope whoever’s reading this will be around to see that! As for a driverless world, where no humans are required on board at all: while Silicon Valley likes to make you believe this scenario is just a heartbeat away, major players in the car industry are saying that this will take 15 years at least.
So like with any revolutionary technology, there is uncertainty regarding the future. Uncertainty breeds rumors, myths and a search for answers – but it also breeds curiosity, hope and excitement. It is exactly for these reasons that 2025AD was created: to start a dialogue, to discuss and take into account fears, reservations and new ideas - and to make sure people are getting a chance to voice their opinion when it comes to automated driving. So go ahead and tell your fellow enthusiasts and decision-makers what you personally think about it – the stage is yours!
Do you have any fears of a fully automated driving future? What is the biggest misconception about Automated Driving that you can think of?