Freightliner's Autonomous Semi-Truck Is Overdue
An overdue technology for a vulnerable industry.
To see just how reasonable a self-driving semi-truck really is, it helps to first understand what an absurd proposition it is in the very first place to ship relatively tiny units of freight vast distances—usually under the watch of some sleepy, caffeine-addled dude. While there are of course good reasons (sometimes, at least) why this particular transportation inefficiency persists in the face of container ships and trains (each capable of hauling hundreds of containers at a time), it can still seem goofy, like digging a mineshaft with a plastic spoon.
As you may have already heard, the truck manufacturer Freightliner has unveiled the first-ever prototype for a semi-truck capable of fully autonomous operation, dubbed the Inspiration—"fueled by the energy of infinite inspiration" (lol). It's already licensed to operate on public highways in Nevada, which two of them are doing as you read this. According to IEEE Spectrum, the acquisition of said license came only after some 16,000 hours of testing. The truck is able to follow GPS directions, obey traffic rules, and stop/slow down/speed up, as dictated by traffic conditions.
This is what's categorized by the DOT as autonomy level 3:
Limited Self-Driving Automation (Level 3): Vehicles at this level of automation enable the driver to cede full control of all safety-critical functions under certain traffic or environmental conditions and in those conditions to rely heavily on the vehicle to watch for changes in those conditions that would require transition back to driver control. The driver is expected to be available for occasional control, but with sufficiently comfortable transition time. The Google car is an example of limited self-driving automation.
So, the driver has to be there and be available, but is free indulge in distractions: watching movies, playing video games, reading, whatever. They won't be called on to suddenly avoid a collision or make split-second decisions, but may have to take control within that "comfortable transition time." It sounds OK, really—what people often incorrectly imagine autopilot to be. Chilling hard on the wide open road, watching Seinfeld DVDs.
Driving a truck as it currently stands is an absolutely ridiculous thing to do. The task is to sit there for many hours at a time mostly just staring and making subtle corrections, which, as anyone that's ever driven a long distance knows, is more or less an automatic process beyond a certain point, like breathing. It just doesn't make very much sense for a person to be doing this.
Driver fatigue is responsible for about 30 percent of all truck-related accidents, with distracted driving not far behind. These accidents wind up injuring some 130,000 people annually, with about 5,000 of those injuries being fatal. 98 percent of all accidents involving heavy trucks involve at least one death (trucks are huge).
Freightliner claims to have found that drivers piloting autonomous trucks were 25 percent less drowsy on average than drivers behind the wheel of conventional trucks. The whole pitch is below:
Road safety is increased by the intelligent networking of the assistance systems. Fuel consumption is reduced due to more uniform traffic flow and powertrain optimization. Vehicle component strain is reduced as a result of the more uniform traffic flow. Maintenance and repair costs are reduced because of less component strain and fewer accidents. Transport logistics are more efficient due to predictive route planning. Driver stress is reduced in monotonous driving situations. Drivers can optimize time with the ability to take over dispatching tasks while on the road. The entire transportation sector's reputation is improved by increased safety, efficiency, reliability and environmental performance.
Long-distance goods transport across the country is by and large dominated by trucking and, arguably, this market share (83 percent in the US, vs. about 6 percent for rail) will only increase thanks to "just in time" manufacturing and the resulting newfound ability to micromanage supply chains efficiently (e.g. make relatively small amounts of something on-demand). Yet, it's still an industry vulnerable to containerization and its natural affinity for globalized markets, security, and automated processes.
Indeed, while trucks own shipping within the US, the global transportation system is dominated by container shipping, with about 90 percent of all dry, non-bulk goods being transported via container and container ships. And containers, once they hit dry land, have a natural connection to rail transport, where a train might simply take over the role of a ship, moving goods to centralized distribution points en masse. Not only are container trains (intermodal trains) more efficient from a fuel perspective by orders of magnitude, it's possible to move hundreds of containers vast distances with two-person crews. (Is it a coincidence that the promo photo above features a railroad bridge in the background?)
In other words, the long-haul truck industry isn't completely safe, and, as such, Freightliner knows it needs to innovate. If anything, it's surprising that auto-trucks took this long.
Are you inspired?