Autonomous tech can cut trucking operational costs by half
Transportation is expensive. Many factors influence this, ranging from poor asset utilization to suboptimal fuel consumption. Today, many fleet managers can hope for at best marginal improvements to efficiency. At the same time, the largest untapped opportunity to reduce transport costs lies with autonomous trucks.
But what is the potential gain from the self-driving tech in trucking? Since there isn’t a single fleet of fully autonomous trucks on the road today, we don’t have an experimentally proven figure. However, we can calculate the potential by analysing operational costs, and making reasonable assumptions as to how they would change.
This extract from a report by the American Transportation Research Institute (ATRI) splits out operational costs of trucking on a per mile basis.
Looking at the 3-year average ending in 2016, we can see that the top cost items are driver compensation (39%), fuel (27%), truck depreciation in the form of lease payments (14%), and maintenance (10%). Let’s look at each item and consider how they could change with self-driving trucks.
Driver Wages
This cost category will cease to exist. No trucker in the cabin means no wages or benefits.
Fuel
Fuel spend is a function of prices and consumption. It is very likely that in the future autonomous trucks and cars will be electric, however that transition will take time — Bloomberg estimates that electric vehicles will account for 54% of new car sales by 2040. Given that this exercise is about autonomous and not electric vehicles I will consider diesel-powered trucks. Nevertheless, it is highly likely that electric trucks will be even cheaper from the energy consumption standpoint.
If we compared a self-driving truck with a manned truck at the same point in time, they’d have the same price per gallon, so the cost per mile would only differ based on consumption. Major levers in reducing diesel consumption are driver behavior, platooning, and idle engine running. According to this article in MIT Technology Review the best drivers are 30% more fuel-efficient than the worst ones. According to Peloton tests, platooning can save on weighted average 7% of consumption for 2 trucks. Taken in conjunction, these factors lead me to conservatively estimate potential fuel savings of 10%.
Truck Lease / Purchase Payments
It gets interesting here, because it is commonly argued that autonomous trucks can drive twice as much as manned trucks, because there is no driver in need of rest, hence they are twice as efficient. This is trivially true, but truck and trailer lease is just 14% of overall operational costs.
An autonomous truck is likely to be more expensive due to additional hardware & software costs. At the same time, there is no need for many elements needed for the human driver, for example sleeper berth or controls. Overall, they’re going to be more expensive (let’s assume 50% more), and twice as efficient. However, we’re talking about 14% of operational costs here, so the impact of better asset utilization on overall costs will not be as large as it may seem.
Repair & Maintenance
On the one hand, this should go up, because the autonomous technology is more complex and high-tech. Replacing LIDAR sensors or malfunctioning cameras will be costly. On the other, with fewer mechanical parts used to control the vehicle by a driver and more efficient driving, the need for repairs goes down. I will leave this item as is.
Insurance
Insurance is used as a collective way to manage risk. The risks associated with driving are mainly road accidents. If we consider a self-driving environment, i.e. when most cars on the road are autonomous, then the risk of accidents should be much lower, or close to zero. According to Morgan Stanley, 95% of all accidents will be avoided with level 5 self-driving technology. The cost of insurance should fall accordingly, so I’ll bring it down to zero.
Permits & Licenses, Tolls
Remain the same. However, there may be significant changes to the way road infrastructure is financed when cars on the roads are mainly self-driving.
Tires
On a per-mile basis this stays the same or goes down slightly due to better driving.
In Conclusion
After summing up all the assumptions I’ve made, it looks like an autonomous fleet is 40–50% cheaper to operate than a regular fleet. This does not include additional benefits that can come from more automation, e.g. auto dispatching, improved fleet visibility, lower overhead, and other factors. Even though automation would benefit a regular fleet, a self-driving fleet will allow for a higher degree of it.
Essentially, there is no doubt that fleets and OEMs will pursue such a large opportunity. Lower transportation costs will create a multiplier effect that will ripple across other industries. For example, it will boost e-commerce. Here is a very peculiar list of areas that self-driving tech will affect by Geoff Nesnow.
