Autonomous Vehicles: The Empty Re-positioning Challenge

Over the next 10 years it is estimated that the number of autonomous vehicles on U.S. roads will become a mainstream force. The rate of progress to this goal is accelerating, building on the success of technology-enabled safety features such as lane assist, adaptive cruise control, and emergency braking. We first saw these technologies come together with self-parking features, and 15 years of continual innovation and evolution are now at a point where self-driving vehicles are a reality.

The stress that will be exerted on already stretched infrastructure will be significant because of the very nature of this leap forward in capabilities. Roads, enforcement, regulations, and the legal system are all designed to work within the framework of a human-controlled automobile. Roads in particular are organized to protect us from ourselves, with speed limits fixed irrespective of the road conditions and flow patterns designed to move a person from A to B.

There is a consensus that car sharing services will continue to expand and with autonomous vehicles go into overdrive as car owners turn loose their vehicles when not in service of the owner. I would not rule this out but it is an overreach to assume that people will do this by default. Personal property is just that and I, for one, would not take kindly to having my vehicles used by strangers unattended. I also don’t think people realize how much the cost of maintenance and depreciation is per road mile, therefore the economic proposition of car sharing may not be a good deal for everyone.

For people leasing vehicles, the mileage restrictions alone prove onerous to car sharing use cases. This is important because lease penetration broadly is about 27% of the total U.S. car market, but as you move up the price tiers the lease penetration increases to 50%. Self-driving technology is an expensive add-on, even more when coupled with EV and hybrid drive systems. It is predictable that over half of all self-driving vehicles will be leased.

So let’s assume that the majority of the market for self-driving vehicles, acquired for personal use, will be strictly for personal use.

Putting aside the obvious challenges that autonomous vehicles present to the legal system – liability – let’s take a look at something not often talked about in the context of autonomous vehicles. Empty repositioning, the process for moving vehicles that do not hold passengers in order to retrieve and transport passengers.

I was first introduced to this problem while looking at an investment in the ocean shipping industry. The flow of trade between countries is rarely equal, the U.S. brings a massive number of full shipping containers from China while sending a much smaller number to China. Left unattended, China would eventually run out of shipping containers, so there exists a well understood process of empty shipping container repositioning. Container ships returning to China carry empty containers in addition to loaded containers. The process is not that complicated, made easier by the fact that shipping containers are standardized in size and configuration and a shipper doesn’t care whether they get a Maersk or APL branded container, they are all the same.

The empty repositioning problem in shipping is solved by loading empty containers on ships headed back to Asia. This is complicated by the fact that empty containers have to be loaded at ports which are not the final destination, but port operators and shipping companies have developed processes to ensure that empty containers are loaded optimally for weight and unloading of full containers at other ports. None of this applies to cars though.

The challenge to public infrastructure comes from the loading of roadways with vehicles that do not contain passengers. For 4 decades the investment in transportation infrastructure has favored incentives to high occupancy vehicles (HOV) over single occupant vehicles (SOV) and public transportation. As HOV lanes, which in CA see about $2.5 billion of investment each year, have proven to not increase carpooling and remain underutilized (LAO report link), toll roads have gained favor.

Unfortunately, car sharing and HOV lane utilization continues to lag projections and the typical HOV lane carries fewer passengers per mile than a non-HOV lane. The challenge that self-driving vehicles presents is enormous because in that scenario a typical SOV will be making 4 trips per day instead of 2.

Autonomous vehicles represent a cataclysmic tipping point for how governments allocate funding for roadways. The entire public transportation system is designed around a premise that increasing congestion will drive people from their cars on to mass transit. In dense regions that are land constrained, the Bay Area being representative, public officials don’t even bother to obscure the strategy increasing the pain of driving.

Here we see a technology that massively disrupts not only traditional businesses, but also public policy on mass transit. Drivers don’t have to drive during their commute, and can send their vehicles home when they are not using them, avoiding the cost of parking. Transportation planners are wholly unprepared for this future, and given the hundreds of billions of dollars that have been spent on public transit systems that are inflexible and capital intensive hub-and-spoke systems, I forecast a lot of pain ahead.