I was reminded this week of a 1983 episode of the classic TV show “Knight Rider” featuring a self-driving truck named Goliath as the villain-of-the-week who battled heroes Michael Knight (played by a young David Hasselhoff) and K.I.T.T., a self-aware and self-driving Pontiac Trans Am. This boyhood memory clanged into my awareness because now, 33 years later, self-driving trucks are racing from science fiction into reality, although none of them sound like William Daniels (who provided the voice of K.I.T.T.).
At Tuesday’s TechfestNW conference in Portland Oregon — where Daimler Trucks North America (DTNA) has its headquarters — two chief engineers shared the company’s ambitious, pragmatic and exciting vision for self-driving trucks in a rare public appearance.
Steve Nadig, DTNA’s Chief Engineer of Mechatronics, and Al Pearson, DTNA’s Chief Engineer of Product Validation, told the story of how Daimler Trucks’ Global CEO Dr. Wolfgang Bernhard let them know in December of 2014 that they would have to demo a semi-autonomous truck by May of 2015 at Hoover Dam near Las Vegas.
With the clock ticking, the two engineers went back to basics, starting with clay models and then shifting to digital models in order to rethink the truck both inside and out.
Nadig and Peason’s team faced more than just engineering challenges: they also had to work with the state government of Nevada to have the truck accumulate 10,000 miles of incident-free driving before it could be licensed.
148 days later, they hit their mark with the Inspiration Truck, a Class 2 semi-autonomous semi-truck (perhaps they should have called it the Semi-Semi?) that answers DTNA’s key questions, “What’s next?” And “How will we drive tomorrow?” in tons of metal instead of mere words.
(Note: a fully autonomous vehicle is Class 5, and a vehicle with no automatic driving capabilities whatsoever is Class 2.)
Humans still behind the wheel, but only paying attention sometimes…
Unlike other vehicle companies that imagine a time when human freight drivers will no longer exist — or perhaps a time when humans are not allowed to drive at all — Pearson doesn’t think that an era with zero human truck drivers will arrive anytime soon.
“Total autonomous is possible now,” Person said. “But only if the environment is controlled. We think it’s going to be a while before we can ensure safety in all conditions.”
With a human driver still in the Daimler truck, that driver’s role will evolve, piloting the truck in the too-complex-for-AI terrain of cities, and then letting the truck drive itself on the highways.
Freed from highway driving, the human driver will manage logistics from the road, pushing that role out from a central or regional office to the edges of the shipping company’s footprint. (More on this later.)
Why develop the Inspiration Truck?
Pearson and Nadig shared a cluster of compelling reasons for self-driving trucks: as the economy grows and becomes evermore global, freight will only continue to be grow alongside it, and self-driving trucks increase freight capacity.
With “platooning” — where one self-driving truck follows another on the highway at a distance of 15 meters or less — the first truck cuts through the wind and gives an aerodynamic benefit to each of the trucks that follow it.
15 meters is too close for a human driver to follow, but with V2V (vehicle to vehicle communications) the self-driving trucks take slower human reaction time out of the equation.
In addition, freight companies would benefit from trucks like the Inspiration because of increased safety features, reduced fuel consumption, reduced vehicle component strain, reduced maintenance and repair, predictive route driving, reduced driver stress, optimized driver time and more generally an improved reputation for the company in question.
Since the Inspiration truck is not striving to eliminate human drivers, Daimler’s approach to self-driving is, in Nadig’s words, “a relatively simple combination of technologies,” that includes an internal camera, two cameras in each mirror (one looking at the blind spot and back down the trailer, a “big safety improvement”), and a radar sensor focused out in front of the truck.
As Nadig observed, using this technology “reduces drivers’ stress and gets them to an optimum level,” but this is only for highway driving: “there are still lots of skills being used on city streets.”
Daimler’s lean approach requires less up-front computation than what other self-driving experiments (like Google, Uber, Toyota, Nvidia) are finding they need, although particularly with V2V and platooning there is still a formidable amount going on under the digital hood of the Inspiration.
As the session shifted into audience Q&A, Nadig observed that “there’s a social aspect” to autonomous vehicles as well as a technological one. “Think about getting on an airplane without a pilot: the technology will go faster than society is prepared to react to it,” Nadig said.
Implications: Daimler’s approach to self-driving trucks is evolutionary rather than revolutionary, which I intend as a compliment. DTNA’s focus is on making its current customers happier and happier by improving the most expensive tool their customers buy and maintain: trucks.
I’m most intrigued by how the Inspiration reimagines — and reengineers — the role of the human driver into half driver and half logistics manager: it’s a rare moment in autonomous vehicle thinking that thinks of humans as a resource worth investing in rather than as a disposable cost center.
One possible outcome to this shift is that drivers also become knowledge workers, higher skilled, higher paid, managing deliveries, inventories and other things at the edge of a freight company’s geographic footprint rather than having things managed mostly in the home office. This could make the freight company more responsive and agile in how it approaches its business.
Moreover, with excess human capacity in the Inspiration, in what other activities could a freight company — or DTNA itself — invest?
Recall that Amazon Web Services (currently the fastest-growing part of Amazon’s business) started as an answer to the question of what to do with the company’s excess web server capacity, and an opportunity for DTNA to do something similar with its own excess capacity in transport and logistics quickly comes into focus (Thomas Friedman discussed UPS doing something similar in his 2005 book The World is Flat, and I’d be eager to see how the larger carrying capacity of DTNA trucks grew that opportunity.)
DTNA’s level 2 approach to self-driving trucks also shows that while the world in which humans no longer drive except for pleasure is still a long way off, we’ll start sharing the roads with non-human drivers in the very immediate future— and not just near Google’s Mountain View campus.
But can I get my car to talk to me in William Daniels’ voice, like K.I.T.T.?