Robotaxis: cost per mile calculations could drive plans off course

Receive free Driverless vehicles updates

We’ll send you a myFT Daily Digest email rounding up the latest Driverless vehicles news every morning.

The traffic transformations promised by robotaxis include a drastic reduction in fatalities, traffic jams and passenger expenses. But as the technology inches closer to regulatory approval, vehicle operating costs remain opaque. 

Rates charged by General Motors’ Cruise robotaxi service in San Francisco are comparable to rideshare services from the likes of Uber. A 10-minute ride costs less than $10 — with a $5 minimum charge plus $0.90 per mile fees and city tax. Waymo, which also operates in the city, cannot yet charge for rides. But it shows passengers an approximation of similar rates. 

In theory, rides should be far cheaper. The Center for Global Policy Solutions believes robotaxis will cost about $0.75 per mile to run, including insurance and maintenance. In 2016, Morgan Stanley estimated that if shared autonomous vehicles become widespread, they would be even cheaper at $0.50 per mile. Lux Research thinks running costs will be $0.42 per mile.

Price comparisons are complicated by variations in location, time and car utilisation. Providers could cut rates as fleets grow. Others, however, may increase fares to recoup R&D costs. 

Autonomous vehicle technology has been in development for close to two decades. In 2004, the US Defense Advanced Research Projects Agency sponsored its first AV competition. Yet costs remain high. Cruise targets annual revenue of $50bn by 2030. But in the first quarter of the year, it reported an ebit loss of $561mn on just $30mn of revenue. 

Add to this expenses such as cleaning. Without the presence of a driver, robotaxi passengers might be more inclined to make a mess. Remote supervision of vehicles to ensure a safe service will raise costs too. This could keep rates in line with rideshare companies. 

There are two ways to lower rates. First, use radar and cameras instead of Lidar (light detection and ranging) systems. They are two-thirds cheaper, though replacing them limits an AV’s ability to recognise objects, which could make them more dangerous. 

The more plausible alternative is to add entertainment systems, paid for by passengers or used to show adverts. If customers can bear that aggravation, robotaxi rides might even be free.