A multitude of other drivers in a stupendous range of other types of vehicles seems easy for us experienced drivers to maneuver around, but imagine trying to program a machine to recognize and respond properly to the entire range of possible interactions on the road. Consider, for instance, how complex operating a heavy, fast-moving vehicle surrounded by other heavy, fast-moving vehicles really is.
There’s the straightforward task of keeping within your lane as you move at a relatively constant speed, but there’s also the not-so-straightforward task of slowing down, or stopping, or swerving around other vehicles in your path. Then there are varying weather conditions, light from many sources coming from nearly every possible angle, crumbling unkempt roads, and forced detours along any predetermined route.
Add in pedestrians, bikers, erratic traffic flows, along with the necessity of maintenance and refueling, and you can understand why the dawn of the Autonomous Vehicle Age still appears some distance over the horizon in late 2020.
Companies like Tesla are tackling the autonomous driving challenge head on. They’re essentially attempting to replace human drivers with automated systems without any significant steps in between. Really, it’s amazing these companies have accomplished as much as they have, with autopilot, lane assist, advanced driver assistance and collision avoidance systems, and on and on. Yet going fully autonomous, as in letting vehicles drive themselves with no human backups on board, remains well out of reach for current technologies.
But what about autonomous vehicles operating in places other than on public roads?
You can think of full autonomy as the big challenge all the major players are attempting to break into a series of smaller steps. We hear mostly about the companies trying to achieve milestones like:
Level 4 autonomy, which means self-driving but with a human on board
Teleoperations, which allows remote humans to take over
Platooning, with one human driver operating a convoy of otherwise autonomous vehicles
But other companies are coming at the issue from an entirely different angle.
Instead of trying to plunge self-driving systems headlong into the complex environments of public roads, companies like Rio Tinto and BHP are setting autonomous vehicles loose within the confines of their mining operations. In these settings, the routes and terrain the vehicles are tasked with navigating are far more predictable, the only other vehicles they encounter are likely operated by similar programming, and the tasks tend to be highly repetitive.
As Perjohan Rosdahl, head of Off-Road at Volvo Autonomous Solutions, explains,
Automation has struggled in the mainstream automotive world because they are trying to get autonomous vehicles to work everywhere and safely coexist with all the variables of life – cars, trucks, bikes, people, dogs, cats – you name it. Solving all these issues at the same time is proving to be an enormously complex challenge, even for the world’s biggest automotive and technology companies. Our approach is to start small, in a tightly confined environment and build on our successes over time. A perfect place to start is quarries, which have clearly defined load-and-dump locations over generally short circuits.
While on the one hand it’s easy to take for granted how complicated a task like driving to work can be until you imagine trying to program all the necessary moves and decisions ahead of time, on the other it’s also easy to overlook how tedious and repetitive some vehicle operators’ jobs can be. And tasks that are tedious and repetitive, like driving trucks back and forth between loading and dumping sites, are ideal candidates for automation.
And another benefit to automation comes into play once the route takes drivers underground – safety. Self-driving trucks are already operating in quarries, and tunnels seem like an obvious next step. Rosdahl explains,
We are starting small with less complex use cases and will build on our successes. With the right customer partners, the next step could be underground mining and tunnel applications – autonomous machines (especially electric ones) work just as well in the dark as in the light, and it’s good to remove people as much as possible from these hazardous locations. From there we could focus on large earthmoving projects that are still contained but have more variables to cope with, as our technology becomes more embedded over time.
Field of Autonomous Dreams
The other proving ground for self-driving vehicles that is already serving as an intermediate stage in the transition to full autonomy is farms. From tractors seeding the fields, to harvesters and grain carts, and even aerial drones scanning strategically for spots to fertilize, the agriculture industry is quickly becoming a hotbed of innovation.
In addition to the benefits enjoyed by mining operators, autonomy also presents an ideal solution to some of the challenges faced by farmers as their industry undergoes momentous transformations. “What I’m beginning to tell farmers,” Scott Shearer, a professor at Ohio State University, says, “is to focus on four or five things they’re really good at and hire the rest.” It turns out farmers are being forced to do the same amount of work – or more – with fewer hands.
“Finding labor has been a big issue,” says Steve Moffitt, the head of D-Dowson Farms in Illinois. Many traditional farming families are finding themselves in the same position of having to do more with less, which makes automated options even more appealing.
“As farmers, we always think we can do it better, cheaper, and faster, but our role is changing,” Moffitt says. “We are starting to wear more of a management hat than a tractor driving hat. This technology allows us to give up things that are out of our efficiency circle.”
As machines’ capabilities are maxed out, the opportunities for increasing productivity inevitably diminish. But whereas a human driver can only sit atop a tractor for so many hours a day, a machine-driven vehicle has no such limitations. Once the total operating hours are thus extended, new opportunities will have to come from better algorithms making better decisions about how to take on the work most efficiently.
Julian Sanchez, Director of Emerging Technology for John Deere, explains,
In the next 10 years, rather than talk about the horsepower on equipment, we’re going to talk about its IQ. Farmers and machinery companies will measure success by how intelligent a vehicle behaves and the contextual decisions it makes in the field without human interference.
And these are precisely the kinds of forces that drive innovation.
So, will the next big innovations on the path to fully autonomous vehicles be made by manufacturers of mining or farming equipment? Or will they come from the efforts of companies like Tesla and Nikola, who are both trying to be first across the finish line with fully autonomous semis? Right now, it’s impossible to say. But one thing is for sure: with all these players attempting to find autonomous solutions to such a diversity of challenges in their own industries, it’s only a matter of time before one of them achieves something truly remarkable. And once one of them makes it across the finish line, the others will be quick to follow.
Follow us on:
You can also sign up for our newsletter so you never miss a post. Just scroll down to the form in the footer.