top of page

Self-driving trucks are still pie in the sky



People tend to overestimate the impact of technology in the short term and underestimate the effect in the longer term. Although the saying is known as Gates’ Law, the original quote is ascribed to Roy Amara, a Stanford computer scientist, somewhere in the 1960s. In general, the broad diffusion of new technologies takes much more time than most expect after they have listened to an overly enthusiastic entrepreneur or have seen it introduced in a pilot on a small scale. Amara’s law is especially true when major investments in infrastructure are required or a wide change in mindset is essential. Self-driving cars are certainly covered by this law and the more so for the complexity of city centres. Bloomberg’s Ira Boudway notes that key players in the industry as Waymo Via, Aurora and TuSimple are now shifting their focus to trucks on highways and are in full swing of building alliances with leading truck producers.


In one of their imagined futures, robot-trucks will soon drive from depot-to-depot, leaving the more complex routes to human drivers. But even that is for now pie in the sky. As Boudway notes, given the immense challenges, self-driving trucks may end up in the same ‘trough of disillusionment’ as robo-taxis are just disappearing. Given their heavy weight, when something goes wrong with trucks, the effects are soon catastrophic. Every other incident will increase opposition and further delay of widespread introduction. “The heart of the problem for driving is predicting about five seconds into the future,” says a former engineering manager for simulation at Aurora, “but with trucks that magic number doubles.” At this stage, only some highly controlled, small-scale experiments are taking place. Before we see self-driving trucks around us in our self-driving cars, massive work has to be done and huge investments are required to build the right infrastructure and systems that keep everything running smoothly and prevent collusions.



bottom of page