There are many questions raised by the recent news that — to cut a long story short — self-driving cars are not going to happen. Top of mind to many SupplyChainBrain readers must be: What about self-driving trucks?

The answer appears to be: They, too, are not going to become a reality any time soon. We should instead turn to the onerous, expensive, but ultimately far more realistic task of reviving America’s neglected rail network, which will achieve exactly what we are trying to get self-driving trucks to do.

It’s time to admit it: driverless trucks are hooey. Take this recent account from the New York Times, about a trial run in March 2022 of a truck developed as a partnership between Kodiak Robotics, a self-driving start-up, and U.S. Xpress. The truck completed deliveries over a distance of 6,300 miles in five days — twice what a solo human truck driver could have achieved. But wait! Because the technology doesn’t actually work, it required a rotating team of specialists in the cab of the truck, so that someone could take control of the vehicle if anything went wrong. Apparently, these “safety drivers” needed to grab the wheel multiple times. 

Read more: Self-Driving Trucks: The Dismal Truth

Even the proposition that a driverless truck could be used exclusively between urban “hubs,” and only on the highway where conditions are more “structured” is revealed as nonsense, because the technology is not yet capable of responding appropriately to “less common situations,” such as a three-car pile-up happening right in front of a truck. Less common these might be, but more than 4,000 people die every year in the U.S. in truck accidents; mostly in passenger cars struck by trucks. 

The idea of a driverless truck, whilst achingly attractive, is the equivalent of claiming the auto pilot can fly the plane without a pilot. It’s a great piece of technology that gives pilots a bathroom or nap break while the hazards in the air are at a negligible minimum, but it can’t do the really tricky stuff we need human brains for, such as take-off, landing and coping quickly with mid-air emergencies. And, by the way, a 12-lane highway presents infinitely greater and more frequent hazards than blue skies at 30,000 feet. There is no single minute of a truck’s long-haul journey that does not present immediate, complex and potentially fatal dangers.

But the idea of sending freight over long distances in an almost completely hands-off way, between cargo hubs close to population concentrations where it can be split up and driven off to final destinations, remains a very good one. Freight companies do this every day, with minimal human intervention, via rail. 

Railroads are the most fuel-efficient way to move freight over land, moving one ton of freight nearly 500 miles per gallon of fuel, on average, and with a total of railroad deaths of 893 in 2021.

The trouble is that, despite constituting the largest rail transport network size of any country in the world, railroads in the U.S. are in a parlous state. 

It doesn’t help that the railroad owners are currently in a death-spiral conflict with their employees over controversial working conditions, while they continue to log profits of around 50% of revenue. Or that relatively clear-cut missions such as, say, to build a high-speed passenger rail link between Los Angeles and San Francisco — which would also benefit freight flows — get hopelessly bogged down in mismanagement and local political horse-trading. (French high-speed-rail expert, SNCF, left the project in disgust in 2011 for Morocco, which opened Africa's first high-speed rail line in 2018.) 

Of the $1.2 trillion Bipartisan Infrastructure Law passed earlier this year, $66 billion is allocated to rail; however, the vast majority of those dollars will go to passenger rail projects. Rail freight operators, including BNSF Railway Company (BNSF), CSX Transportation, Inc. (CSX), Norfolk Southern Railway (NSR), and Union Pacific Railroad (UP), are left to maintain their own infrastructure. The Association of American Railroads trumpets private investments from 1980 to 2021of nearly $760 billion — well above $20 billion a year — on capital expenditures and maintenance expenses. But problems with rail freight service have become so acute that, in April, the Surface Transportation Board ordered the four biggest railroads to develop service recovery plans and provide monthly information about their progress. 

It would appear rail freight in the U.S. needs an urgent rethink, including potential nationalization. After all, the nation’s highways are in public hands. Why not the equally critical rail infrastructure?

Aside from the fantastic ability to make split-second decisions in highly complex situations that continue to beat any computer, humans also have a tendency towards wishful thinking when it comes to the abilities of new technology. History is littered with examples, but my favorite is the Mechanical Turk. This was a fraudulent chess-playing machine constructed in the late 18th century to impress Empress Maria Theresa of Austria. It appeared to be able to play a strong game of chess against a human opponent, and managed to beat famous challengers such as Napoleon Bonaparte and Benjamin Franklin. Finally, it was revealed to be a hoax – there was always a human chess master hidden inside. 

But, such was the credulousness of those who examined it, it took 84 years for the deception to be revealed. Let’s cut to the punchline with driverless trucks as a solution to our ever-burgeoning long-distance freight needs, and focus now instead on rail.