What’s holding back the self-driving car revolution?

“The 360” shows you diverse perspectives on the day’s top stories and debates.

What’s happening

According to promises made a few years ago by automakers and the tech industry, we should all be sharing the road with millions of driverless cars by now. In 2016, a long list of companies asserted that within five years the once-unimaginable dream of fully self-driving cars would become a normal part of everyday life.

That hasn’t happened, of course. Major advances have been made, but even optimistic forecasts suggest it will be years — if not much longer — before regular consumers will be able to comfortably put their attention on anything other than the road while making their commute.

Right now, a number of car companies offer what are often called driver assistance features that can automatically handle a range of driving tasks like cruising, changing lanes and braking. This is what’s known in the industry as Level 2 automation. But each of these systems, even ones like Tesla’s that have been labeled “full self-driving,” requires a person behind the wheel who must maintain focus as if they were driving, so they can intervene if needed.

In a handful of cities, a few companies are testing driverless vehicles that can navigate without any human control. Waymo, an autonomous vehicle firm owned by Google’s parent company, offers a driverless taxi service in parts of Arizona. Those vehicles use substantially more hardware than the typical car with a driver-assist system and only travel on roads that have been previously mapped by human drivers.

The desire for fully self-driving cars isn’t just about convenience. Experts see autonomous vehicles as a means of reducing the enormous human cost associated with traditional driving —in particular, the 1.3 million estimated deaths worldwide per year caused by traffic accidents. The technology could also reduce congestion and speed up commutes, lowering emissions that contribute to climate change.

Why there’s debate

The simplest reason that the dream of autonomous cars hasn’t been realized is that driving is much more complex and difficult to replicate than automakers anticipated. Experts argue that, for all of the mistakes humans make on the road, our brains are equipped to handle unexpected situations — like an emergency vehicle parked in an unusual spot or a sudden change in road conditions — in a way that artificial intelligence struggles to mimic.

Others lay some blame on carmakers for rolling untested self-driving features onto the road and making lofty claims that prompt drivers to push beyond their vehicle’s capabilities. Misuse of self-driving features has caused deadly crashes, which in turn contributed to deep skepticism about the safety of self-driving cars among the public. Some proponents of self-driving technology, on the other hand, argue that authorities and critics are holding autonomous vehicles to an unfair safety standard relative to what’s expected of human drivers, which they believe limits how much companies are able to refine their products through real-world testing.

Another issue may be humans themselves. There’s some research to suggest that vehicle automation has stalled at Level 2 because our brains aren’t wired to maintain attention when not directly participating in an activity, which is what many driver-assist systems ask us to do. Finally, there are some experts who believe that the task of creating cars that can navigate every imaginable road scenario may simply be impossible.

What’s next

There is bound to be continued experimentation, possibly limited by future laws about where and under what circumstances this technology can be deployed. Transportation Secretary Pete Buttigieg said in March that he expects “meaningful” federal regulations to be put in place in coming years.

Perspectives

Replicating human driving is much more complicated than automakers anticipated

“The community overestimated the potential of even the most advanced technology and underestimated the capabilities of even the least trained human driver. Driving, despite what many think, is a complex, dynamic effort at multitasking. Maintaining the speed and position of a vehicle no matter the changes in weather, traffic, road conditions, and the diverse mental, perceptual and motor capabilities of the human driver is not easy.” — Francesco Biondi, Conversation

Unachievable safety expectations are preventing real progress

“No matter how much they improve, theoretically, there will always be a situation that an AI won’t be able to detect and react to. Unless the industry and public agree to accept a flawed self-driving system — one capable of failure — autonomous vehicles on our streets will never become mainstream. Achieving perfection here can’t, and shouldn’t, be the goal.” — Jurica Dujmovic, MarketWatch