Public trust, not technology, is barrier to self-driving trucks

Updated Nov 29, 2013

Truck drivers shouldn’t worry too much about losing their jobs to robots just yet, according to at least one truck safety technology expert. Fully autonomous trucks are several technology leaps away, but the first incremental steps have already been taken and they’re in commercial use today.

Technology already can control truck speed and braking. Why not steering?Technology already can control truck speed and braking. Why not steering?

Following up on last week’s House hearing, which concluded that self-driving cars are coming but no one knows when, CCJ got in touch with Fred Andersky, director of government and industry affairs for Bendix Commercial Vehicle Systems. Andersky was in Washington for the discussion because Bendix, along with Meritor WABCO, is a market leader in developing advanced safety systems for commercial vehicles. Both companies have a keen interest in policy decisions that impact their business.

Though trucks were touched on briefly by the panel, I passed along a previous blog discussing the possibility that driverless technology might come to commercial vehicles ahead of passenger cars, based on the economics: Trucking companies, if the technology does indeed improve safety and efficiency, should be more than willing and able to justify the investment.

Indeed, a fleet of self-driving, giant mining trucks is already deployed in a closed-loop at an Australian site. Most of the jobs done by drivers will be replaced by computers and remote operators by 2015, according to the report from The Australian.

Andersky doesn’t see autonomous vehicles developing in the same way for highway use, though the financial incentive will be there. As was discussed during the hearing, public acceptance is a critical part of the bigger picture.

“Developing the technology is not usually the issue, it’s always the application of the technology where the devil finds its way into the details,” Andersky says. “It’s going to take a cultural shift to get to the acceptance of autonomous vehicles, especially autonomous trucks out on the road. Coming up through the automotive side is going to drive the ability to accept these vehicles.”

He points to public backlash over periodic efforts to increase truck size and weight as an example of how perception of the industry, informed or not, influences policy.

In short, people won’t accept self-driving trucks until they’ve learned to trust their self-driving cars. And that trust will be built through intermediate steps.

Automatic transmissions are an obvious example of a development that handed over a degree of control from the driver to the vehicle itself. (Curiously, it’s also an area where trucking lags the near-universal automotive implementation by decades.)

Likewise, anti-lock brakes, traction control and electronic stability control systems are now accepted with hardly a second thought – and all employ sensors and microprocessors to do the thinking and reacting faster than a driver a can.

Forward-looking, sensor-based systems such as adaptive cruise control and collision mitigation systems are currently in wide use, though not yet standard equipment in either cars or trucks – but that day is not so far away.

To use stability control as an example, the aha-moment comes when the driver acknowledges the system does a better job of negotiating a familiar, tricky turn than the driver had been doing. More commonly, a driver becomes a believer after a sudden swerve to avoid an object in the road “and the system saves his bacon,” Andersky adds.

In turn, collision mitigation systems are more autonomous yet.

Of note, the trucking industry press knows Andersky from Bendix demonstrations, for which he rides his motorcycle ahead of 18-wheelers coming to a stop behind him with no driver intervention.

“I really believe in the technology and how it works. But I also know there’s a driver in the truck if it doesn’t work,” he says, laughing. “I’m not sure how thrilled I’d be if I didn’t know there was at least a fail-safe. We’re not going to be getting rid of drivers as quickly as people might like to think we are.”

So completely self-driving vehicles are years and years away from being common – although the technology is tested on American streets every day. What about the next intermediate step?

The GM representative at the House hearing mentioned the company’s “super-cruise control,” technology that will offer limited, automated steering.

“When you get to the point that the driver can step away from the controls, even only for brief period, that is going to be a major shift in terms of agency thinking. It’s not going to be enough for Bendix or WABCO, or even Google, to say ‘we’ve got it, it works great.’ Regulators are going to do a lot of testing beyond that, in their own way and in their own fashion,” Andersky says. “The technology may be ready in Year One, but it will still probably take a few years for there to be a regulation that moves it forward. Fleets may still be able to purchase that technology, but a lot of that is going to depend on what it does.”

The bottom line: If regulators see a safety benefit, they will approve new technology.

“I don’t think they’ll adapt as quickly as the industry can provide the technology, but they are doing some of the right things in terms of evaluating different types of technology and keeping abreast of what is going on,” he says.

And while Andersky declined to discuss specific Bendix technology or coming products along these lines, he offers a vague timeline: “I don’t think we’ll see anything that we’re talking to customers about for at least 7 to 10 years – and that might be optimistic.”

Still, Andersky’s given a lot of though to what trucking might look like – if and when driverless technology takes hold. But that’s a topic for another discussion.