The public image of autonomous vehicles is a car threading through a busy city, negotiating pedestrians, cyclists and chaos. That is the hardest version of the problem, and it is not where driverless technology is making its steadiest commercial progress. Freight is, and specifically the middle-mile: the long, monotonous highway haul between distribution hubs on the edges of cities. It is the part of trucking that is most amenable to automation precisely because it is the most boring — a divided highway, predictable lanes, no pedestrians, hours of steady cruising — and it is also the part the industry struggles hardest to staff.
The labour dimension is what gives the technology its pull. Long-haul driving keeps a person away from home for days or weeks at a time, the work is gruelling, and the turnover and shortage of drivers willing to do it are chronic and worsening. Automating the highway stretch does not straightforwardly eliminate the job so much as reshape it: the emerging model is a hub-to-hub relay in which an autonomous system handles the long, dull highway middle while human drivers take over for the complicated first and last miles — the depot, the city streets, the loading dock — where judgement, manoeuvring and human interaction still decisively beat any machine. The person moves to the hard, local, human end of the trip; the machine takes the tedium in between.
The engineering is genuinely more tractable here than in the city, but "more tractable" is not "solved", and the honest state of the field is careful expansion rather than a finished product. Highway driving still throws hard problems — a stopped vehicle over a crest, debris in the lane, sudden weather, a construction zone, the sheer physics of stopping a heavily loaded truck — and the consequences of a failure at highway speed and weight are severe enough that the safety case has to be overwhelming before scale is justified. So deployment has been incremental: specific, well-mapped corridors, favourable conditions, extensive monitoring, and a great deal of caution about the transition from a safety driver present to genuinely unattended operation.
For the logistics industry the shift, when it matures, changes the shape of the network rather than simply removing a cost. A relay model built around automated highway legs favours hubs positioned at the edges of urban areas, changes how loads are scheduled and handed off, and rebalances where human drivers are needed toward the local and regional. None of it happens overnight, and the timelines have a long history of being optimistic. But the direction is clear and it follows the same logic as every other successful robot: automate the narrow, repetitive, hard-to-staff task first — here, the dull highway middle — and leave the messy, judgement-heavy ends to the people who are still far better at them.