When people talk about the jobs that are likely to be automated in the coming years, trucking is one of the most common to be cited, with innovations by companies like Otto used to exemplify the risk drivers face.
It’s a risk that a new study suggests is significantly overblown, and the authors provide three reasons why this is so.
A more expansive role than we think
The first reason is that trucking involves quite a bit more than ‘just’ driving the rig, so even if the driving part can be safely automated, that’s not to say that the role of the driver has been negated. Drivers also do tasks such as securing cargo, maintaining logs and performing customer service, as well as obviously loading and unloading the cargo on their truck. There are tools to support some of these roles, but nothing to actually automate them.
Full automation is not around the corner
The second reason revolves around the evolution of the technology itself. As with so much of the coverage related to AI and automation, the hype train has somewhat run away from itself, and the authors argue that safe and efficient automated truck driving is still a little way off.
They remind us that only very limited demonstration events have come close to the level 5 (full automation) standard determined by the Society of Automotive Engineers, and that these are often confused with the much more widespread adoption of level 2 or 3 technologies that provide partial or conditional automation, which are more common.
The researchers note that level 5 technology is not really being tested in practice yet, so based their employment conditions on level 4 automation, which assumes the technology can control the vehicle and monitor some of the operating conditions of the vehicle.
With some companies actively developing tech at this level, it seems a reasonable place to start testing assumptions. The majority of development to date has been on long-haul trucking rather than short haul or local trucking.
Around 1/4 trucks are used for long-haul journeys, with the majority used for short journeys of under 50 miles. This would still place around 450,000 drivers at risk, but it’s less than the millions often projected. It’s also highly likely that adoption across the sector will be staggered rather than widespread, which the authors believe will further reduce the number of at risk drivers, at least in the short to medium term.
Truck drivers aren’t as prevalent as we think
The last factor they believe skews our perceptions is around the number of drivers people think operate in the United States. Most articles that discuss the topic put the figure at around 3 million, but the reality is a much smaller number.
The 3 million figure is often derived by using the government’s Standard Occupational Classification (SOC)’s category of ‘Drivers/Sales Workers and Truck Drivers’, which can include everything from local couriers to delivery van drivers. In other words, not the type of long-haul driver that is most at risk of automation.
When looking purely at those requiring a Commercial Driver’s License, the number falls to around 1.8 million, which even when self-employed drivers are added to bring it up to 2.4 million or so, is still a fair bit below the figure usually cited. But the long haul proportion of this group is estimated to be around 450,000, or 0.3% of the total U.S. workforce and roughly equivalent to the number of welders in operation.
It perhaps also goes without saying that there are numerous regulatory hurdles for companies to overcome before level 4 technology is rolled out en masse. For instance, the infrastructure for repairing broken down trucks is largely missing, much less preventing potential theft of cargo on stricken vehicles. Trucking long distances is also often a complex affair involving multiple parties, especially when crossing borders that may require customs checks.
This panoply of issues is likely to slow down the march of driverless trucks, and while it’s a likely inevitability that the technology will arrive eventually, it gives ample breathing space for drivers to adapt to the changes that are looming on the horizon.
