A self-driving future is farther away than we think, according to Nissan technology boss, who suggested more of a human touch could be what's needed.
- shares

“[Humans] are the best autonomous systems I know of.”

So says Dr Maarten Sierhuis, chief technology director for Nissan, and the man charged with guiding the manufacturer’s autonomous driving technology into the future.

But the NASA veteran, who oversaw and designed autonomous technologies for the US space agency, has joined a growing number of voices in believing fully autonomous vehicles may never happen, certainly not in the way they've been touted in recent years.

Instead, he and his team are now working on a system he calls ‘human-in-the-loop’, where self-driving vehicles will always have an operator overseeing the process and reacting to unforeseen circumstances where a human touch could offer a better solution than an algorithm.

“Why do people think we can have millions of these vehicles just driving around and not needing any human interaction?” posed Sierhuis.

Sierhuis used the aviation industry as an example, where air traffic controllers offer information and guidance remotely, ensuring the safety of aircraft and their occupants.

“We look at air traffic control. We have two pilots in the cockpit and we still need air traffic controllers. We have operators in a control centre who have geographic areas they are responsible for. They communicate with the pilots in the cockpit and make sure everything goes well, especially during take-off and landing.”

Sierhuis’s – and Nissan’s – proposed model relies on a central control hub manned by human operators. Sierhuis stressed, however, these operators wouldn't be driving the cars, only helping make decisions when faced with obstacles outside the autonomous software's scope. In other words, offering that inimitable human trait, instinct.

“In the beginning people said I was crazy,” said Sierhuis. “You’re here to develop autonomous systems. But then you show them examples.

“What is the system going to do when it has to break rules? Are you going to allow it break rules? But how are you going to define what rule it can break when and how?”

Enter Nissan’s Seamless Autonomous Mobility (SAM), a system manned by remote operators who could oversee as many as 50 cars and intervene when unpredictable situations like roadworks, or unmapped areas like carparks, confuse the autonomous systems.

Despite this human touch, Sierhuis stressed autonomous vehicles will continue to drive themselves, with operators stepping in as required.

He offered an isolated example, where a car approached an undivided two-way road blocked by roadworks. With double solid lines dividing the road, human drivers can interact with the road workers, who wave them through when it's safe.

A fully autonomous vehicle, programmed to not break road laws and unable to interact with humans, is forced to remain stationary behind the blockage. That’s where Sierhuis’s human-in-the-loop comes into a play, with a remote operator able to tell the vehicle to cross the double-yellow lane and continue on its pre-programmed journey.

“The system is always autonomous,” he said. “The car is responsible for its safety, is responsible for driving. The human-in-the-loop is supervising and supporting the car in decisions where it needs help by human intelligence.

“This is what we have developed together with NASA. We worked together with a partnership between NASA and Nissan in the United States that I helped set up. We used software that NASA developed for robots on Mars.”

Sierhuis believes the technology would suit large-scale commercial applications rather than individual car buyers.

“What it really is about, what we are embarking upon – to make driverless technology available as robo-taxi services, and as delivery services – is to really move from what I call individual autonomous systems, every car on the road being its own data centre with AI technology in the car, to a system where there is this holistic autonomous system, with intelligence on the vehicle, intelligence in the cloud and human intelligence where it’s needed," he explained.

"And the trick is now, how we can scale to one person managing a whole fleet of cars and providing a safe and exciting mobility service.”

He cited the high costs associated with the system as cost-prohibitive for individual consumers.

“I do think [commercial] is the first application you will see,” he said. “Because the price of the system, the cost is going to be prohibitive in the beginning for consumers to pay. I think the market is just not there to do that.”

He remained cautious about predicting a completely autonomous future, a sentiment being echoed by self-driving car industry leader, Waymo CEO John Krafcik, who said late last year truly autonomous cars with Level 5 capabilities “will never exist”.

Additionally, people still fear self-driving cars. Last week, an American Automobile Association survey revealed 71 per cent of respondents would be too scared to ride in a self-driving vehicle. That percentage has grown from 63 per cent, largely attributed to the fatality in Arizona last year where a self-driving Uber (with a human behind the wheel) killed a pedestrian who walked into the path of the vehicle.

That’s why Sierhuis believes in what he calls “socially acceptable autonomy”.

“That is the key to where we are going. Because building an autonomous system that doesn’t interact with humans, that seems very simple. It seems that we can do that right now with robots.

“But when we have cars driving, we really have to move to an inter-dependant relationship between the human and the autonomous system. That is because humans are social. I always say we are the best autonomous system that I know of.

“We interact and we communicate. And we need to do that with the vehicle, inside the vehicle, we need to provide services to the people inside the vehicle when there is no driver. But it’s also important for people outside the vehicle; other pedestrians, other road users, drivers, cyclists.

“So there is where my path goes about humans in the loop. This is a very important part of autonomous vehicles, and not just with autonomous vehicles, with any autonomous systems we will develop in the future.”

Sierhuis summed up succinctly the future of autonomous vehicle.

“Show me an autonomous system without a human in the loop and I will show you a useless system.”

The future, as it’s been sold to the public in recent years, is still a long way off, it seems.