When an Uber autonomous vehicle struck and killed a pedestrian in Arizona on March 18, most companies put their self-driving tests on hold. The avalanche of press releases revealed just how many different companies are already testing self-driving cars. The list includes Apple, Google, Tesla, nearly every major automaker, computer chip-makers like Intel, Nvidia, and Qualcomm, and a slew of lesser-known technology companies like AImotive, Aptiv, Argo, Aurora, Keolis, Navya and Torc.
Some might wonder what Uber is doing among the ranks of such high-powered engineering firms. After all, Uber goes to great lengths to argue that it is merely a ride-sharing platform that links passengers with independent drivers, not a taxi service that offers mobility solutions. If Uber does ultimately succeed in developing self-driving cars, who will be its independent service providers? The robots themselves?
That may be an existential question for Uber, but it won’t stop the company from exploiting its privileged position as the app that owns the ride-hailing customer. When people want a car to pick them up and take them somewhere, they call (OK, click) Uber, not Aptiv or Keolis. Unless, of course, they’re in China. In China, one app and one app alone owns the car hailing customer, and that app is DiDi.
Get me a DiDi
Beijing-based Didi Chuxing (literally “Beep-beep Travel”), or DiDi for short, is China’s leading Uber clone. It is everything Uber is, and then some. With investors like Alibaba (owner of Alipay) and Tencent (owner of WeChat), DiDi is the overwhelmingly dominant ride-hailing app in China. In 2017 DiDi sold an incredible 7.43 billion rides, nearly twice as many as Uber. Even deep-pocketed Uber gave up the chase and traded its China business for a 17.7% stake in DiDi in 2016.
Of course, DiDi is also developing autonomous driverless cars. Its technology partner is an electric car startup called CHJ Auto. Together they are working to create an all-electric robotaxi that can shuttle passengers around China’s crowded megacities with no drivers at all. But that doesn’t mean an existential crisis for DiDi: unlike Uber, DiDi doesn’t have to pretend to be a mere sharing app. China has regularized the status of ride-hailing services so they are regulated more or less like conventional taxis.
Chinese Internet giants (and DiDi investors) Alibaba, Baidu, and Tencent all have self-driving vehicle programs of their own, but as with Uber in the United States, DiDi owns the customers — at least for now. A new round of competition may be just over the horizon. But it will be tough for anyone to dislodge DiDi, which like Uber in English has simply become the Chinese word for an on-demand car (as in “get me a DiDi”).
Robo-taxis to the rescue
DiDi’s move into autonomous vehicles is coming just as China is moving toward the creation of an all-electric vehicle ecosystem, at least in its major east coast cities. Two huge problems have historically plagued all efforts to roll out true battery-electric vehicles (BEVs): charging and range. People can charge at home, but it’s hard to find a public charging pile when you’re out on the road. Even if you can find one, it can take at least half an hour to get a decent charge.
The electric robo-taxi solves these problems. Each autonomous BEV in the DiDi robo-taxi fleet will continuously monitor its own battery levels. When a robo-taxi has to charge, it will just take itself out of service for an hour. With some pretty basic AI, the fleet can plan its charging around peak user demand. And the range problem disappears as well, since your DiDi ride hail will only be answered by a robo-taxi with sufficient charge to get you to your destination.
Rail plus robo-taxi
China is an ideal environment for the development of robo-taxi ride-hailing because most car travel in China consists of short trips at low speed. For longer trips, people can switch to China’s 25,000-kilometer (15,500 mile) high speed rail network. High-speed rail already connects most neighboring cities in China, with a full national network of 38,000 kilometers (24,000 miles) planned for completion in 2025.
The perennial pitfall of high-speed rail is the “last mile” problem: it may take less than two hours to travel the 285 kilometers (175 miles) from Shanghai to Yiwu, but how do you get from Yiwu’s train station to its famous wholesale market? With future DiDi, your robo-taxi awaits. Everyone’s does. They will be booked as part of the rail ticket itself, and the app will tell you where to exit the station to jump straight into your pre-programmed ride.
No more airport parking
The combination of rail plus robo-taxi won’t just change China. Consider the long-awaited Texas Central rail project for a bullet train to connect Dallas and Houston. The biggest argument against the project will disappear once Dallas and Houston switch from personal commuting via pickup truck to robo-commuting via electric taxi. And even if Texans keep their V6 extended cabs for local trips, they’ll dump them for trips to the airport and the train station. Ubiquitous robo-taxis will make airport parking obsolete.
The simple fact is that BEVs are already cheaper than conventional cars once running costs are factored in. When they’re run continuously, like taxis are, the BEV cost advantage is decisive. The only problems are charging and range, and the fleet solution takes care of that.
If you drive one car, it’s likely to have a conventional or hybrid engine. But in the future, if you access a fleet of continuously operating self-driving cars, they will all be electric. DiDi has a good shot at getting there first, and getting there big. It may be able to solve Uber’s existential question — and in the process become the world’s preferred robotaxi service.