Autonomous traffic is coming – but not in the way you expected

Driverless cars are touted as one of the big systemic innovations of the 21st century. But could this supposed megatrend be blinding us to where the real change is going to come from?

February 13, 2017

Looking at the amount of publicity the driverless car it getting, it is easy to get the impression that the automotive industry is putting all its bets on autonomous cars as the future of personal transport. All the main car OEMs, plus a growing group of new entrants, with Google perhaps the best known (and who knows, maybe Apple, maybe not), are both investing in the required technologies, and eagerly showing the results of these investments.

The motivations of the various actors may not be equally evident, and expending some thought on these reveals a more varied picture.

The struggle to find a business case

The technology roadmap towards driverless cars is for the most part already known. The same cannot be said about the arguably more important business roadmap.

The driverless car, together with many other factors relating to the general digitalization development, will disrupt the current business models of incumbent OEMs, introducing new types of actors and new types of services that do not participate in the car industry  today. Incumbents are  quite understandably afraid of losing their position at the top of a very important value chain, one that has dominated western economies for decades,and having to settle instead for the role of a supplier.

Yet, being viewed as passive in this field would certainly hurt any car company’s position in the stock markets. Thus, for the automotive industry there is no other option than try talk up their game, and continue investing in driverless cars. Whether the intention is really to mass produce autonomous vehicles by 2021, as Ford has said, needs treating with some skepticism.


For the new entrants, the situation is different and varies a lot.

For Google, the driverless car looks like another major platform for its information services and advertising business. From this point of view, its investments in the technology look to be motivated by speeding up the opening of the market, rather than an attempt at becoming a car company.

For the other big player in this field, Uber, this is a question of life and death. Its current business is making huge losses (according to some estimates, approaching $3 B last year). The driverless car looks like the main solution to the problem, as it can decrease the costs of providing rides. Uber currently allows its drivers to retain 70-80% of the customer payment.

To put it bluntly, though, regardless of the big investments nobody has more than guesses about how to make money in the driverless car domain, or when the money can be made, and how much money there is to be made.

The alternative autonomous vehicle

However, there is a domain where at least some more light can be shed on these issues: goods transport and logistics.

Goods transport has the advantage of less severe technological requirements than cars. And it is part of an overall logistics chain that is exploring the benefits of AI.

The main modes of goods transport are trucks, trains, ships and airplanes. Below I will focus on trucks and ships. Most of the arguments apply also to trains and airplanes, in some cases self-evidently.

Taking the last issue first, in goods traffic most of the distances covered are driven along predetermined routes. Typically, only the last kilometers may require ad hoc routing, for example in distributing goods to retail locations. Thus, it becomes feasible to install special equipment supporting autonomy, and set aside given lanes for driverless trucks, as is already being considered in “platooning” schemes.

There are  also fewer people moving around in trucking routes compared to urban traffic.


Shipping routes are also well defined, at least in congested waters, and closer to shorelines and ports. There are now efforts to create an innovation ecosystem in maritime and apply artificial intelligence to a number of maritime tasks.

My own organisation DIMECC, together with major players in the maritime industry, has started that ecosystem activity, aimed at realising commercial autonomous shipping in the Baltic by the year 2025.

Evidently, ships are much more vulnerable than trucks to natural and weather conditions. Nevertheless, even in this case the environment is not as dynamic and unpredictable as in urban traffic. The required intelligence may not pose the same level of risk to life.

The patterns of shipping and logistics traffic result in a less dynamic and more predictable environment, and therefore less severe requirements for driverless trucks and ships than for driverless cars.

On a side note there is also the option to make more use of drones to complement these developments.

The more obvious business case for logistics innovation

From the business point of view, driverless trucks offer several advantages compared to human driven trucks. The lower operating costs, increased reliability, and added efficiency (more time on the road) all add up to major cost savings, and results in higher profit margins.

The business case is clear even with this. On top of this comes the potential of new kinds of services and revenue streams, not possible with human driven trucks.

The same arguments of lower costs, better reliability and added efficiency apply also to autonomous ships. Again, major cost savings can be achieved through autonomy. This is very significant in an industry, where fuel cost savings in the low single digit percentage point range are considered good. Even semi-autonomous systems will help huge efficiencies in routing away from storms or finding more favourable currents.

Autonomous goods transport already has a business logic that can be realised without major changes in business architecture.

Mostly, the current players can take advantage of the cost savings opportunities, as soon as the required technology is mature enough. And, as explained above, this should happen sooner than the maturation of driverless cars for personal transport.

At this point it also becomes clear that autonomy will have a major impact on the whole logistics business and value chain. The cost savings business logic drives the early phases of the disruption, as this is easy to justify, and does not (yet) require major changes in the business architecture. Due to this, the logistics business will be the first to adopt the new autonomous technologies, and actually make money with them.

Once the technology is sufficiently widely adopted, and the benefits of the cost savings fully utilised, the second phase of true value chain disruption will begin. Also in this respect, the logistics business will be the leader, but this is another story.


Instead of following the herd and aiming at the fashionable but blurry autonomous car market, it would make lots of sense for companies to spend serious thought on the clear and present opportunities offered by the goods transport and logistics domain. There are also opportunities for researchers, as there are major open questions that require focused research.

Goods transport is the main use case for autonomy. Let’s make it happen.