For years, driverless cars were in the research and development areas of companies like Google and Tesla. They are just now transitioning out of R&D and into operational or engineering developments.
There are still significant limitations with broad deployment of self-driving cars -- Uber's new "driverless" cars in Pittsburgh
have humans behind the wheel -- but they are approaching a level of operational maturity that enables them to address a huge problem
: the 100 fatalities and 10,000 injuries a day caused by car accidents in the U.S.
The problem is that while there is a good case for driverless cars and planes, it is harder to justify a need for driverless trucks, especially in urban environments, where their operations are impaired by complex traffic patterns, single-lane streets with many obstacles on either side and pedestrians who are unpredictable.
And nowhere is this mismatch between the technology "solution" and the "problem" at hand more obvious than with the recent announcement
by Uber of driverless beer trucks.
To be sure, automated beer delivery makes for a catchy headline, but safety demands that we take a closer look. Unlike cars, driverless trucks have undergone only limited testing -- under ideal conditions, and for relatively few miles. This is important, because the artificial intelligence behind driverless trucks is proving more difficult to develop than similar software for driverless cars.
Uber's beer truck may be a good start, but no company should expect that much of their long-haul trucking fleet will be driverless in 5-10 years. Truck aerodynamics
are more complex than cars because of their size and the variety of loads they can carry. Cars carry humans and a few small items. Trucks carry everything from a load of beer to a load of live cattle.
In short, the work to develop driverless trucks that can be reliably used by companies is far less along than driverless cars. There is no guarantee that the outcome of this development work, and its costs, will lead to commercially useful trucks that are deployed in large numbers.
But there's a superior alternative to driverless trucks. The research has been done, and the operations are already mature.
Instead of trucks, Uber and Anheuser-Busch InBev should be investing for the future in aerial drones.
These drones can fly industrial loads from a terminus to a delivery location in a path that is more direct than existing highways. Though examples of this utility for drones are still few and far between, they do exist across the globe.
Small drones in Australia and Rwanda
are carrying loads of several hundred kilograms. Amazon
has announced it will test drone deliveries in partnership with the UK government. And an Israeli company has a drone prototype
that can carry 1,100 pounds for 31 miles.
Companies looking to use industrial drones can learn from military and intelligence organizations, which have conducted mature operations for large drones that maneuver in adverse situations. Industrial drones can reuse much of the software and operational best practices these organizations have developed over the past two decades.
Scaling up to widely used industrial drones means companies could start replacing trucks on the roads within five years, with enormous benefits to profits, workers and consumers alike.
In addition to being at a more advanced and practical stage of development, industrial drones have key advantages over driverless trucks. First, they do not have to be pilotless. This simplifies their operations because a ready-made population of those pilots already exists: The skill required to fly an industrial drone would be similar to drones already in use by the military and intelligence communities.
Industrial drones can carry loads along a more direct route, as their mission is simply to move a load from point A to point B. An industrial drone can pick up a load at a terminus and deliver it to its location in a straight line instead of following the U.S highway and interstate systems. Even better, an industrial drone is not subject to the limitations that many trucks have today on length, width, height and weight.
Best of all, the operational maturity of an industrial drone brings existing truck drivers into the pool of potential pilots as well. Instead of having their jobs eliminated, truck drivers could be trained to sit in a remote location and operate industrial drones in any part of the country. Unlike driverless trucks, industrial drones could offer jobs to Teamsters in a fourth industrial revolution.
Since these pilots should be able to handle more than one drone at a time, industrial drones would provide efficiencies to customers, improving consumer experience.
We need to reprioritize our investments to improve the trucking industry. Most of these funds should be repurposed to embrace the future with industrial drones while continuing with driverless truck development at a lower priority. Right now, about 10 million trucks
are on the road at any given time in the U.S., posing both safety risks and huge costs. The most-used freight corridors, consisting of 26,000 miles of highways, account for over 95% of that total. The infrastructure associated with these freight corridors require billions of dollars annually to fix and maintain.
Replacing these with driverless trucks does not address this situation in a positive way. Removing trucks from the road and replacing them with industrial drones will lead us into the future with a better safety record and less stress on our infrastructure.