Kyoto: Uber, Doshisha University, and more golf carts

Whenever visitors come to visit us in Japan, we find they are inevitably travelling to Kyoto for a few days as part of the trip. But beyond its myriad temples, illustrious history and tourist hordes, in recent years Kyoto has been quietly accelerating the number of automated driving tests across the prefecture.

Already back in 2015, Kyoto’s Doshisha University Mobility Center tied up with neighbouring Kanazawa University and Nagoya University to kick off a collaborative intelligent transportation systems concept geared towards providing up-to-date maps to automated vehicles. Tests were performed on May 21-22, 2015 at the Doshisha Campus using the University’s test vehicles, which look from the photos like a modified Toyota COMS vehicle kitted out with the regular LIDAR/camera rig. Test vehicles from Kanazawa and Nagoya University were also used in the test, typically a Prius platform with added sensor rig fixed to the roof. And while the goal to demonstrate the ability for the vehicle to drive itself along a defined route while leveraging a local dynamic map reference was successful, there were no plans announced to take this to an uncontrolled environment such as downtown Kyoto where vehicles, tourists and locals mill about the vehicle on a realtime basis.

In November of 2016, real world tests were announced by the Ministry of Land and Transport to take place in Nantan City (pop: 32,770) and Kizukawa City (pop: 74,050). The test platform was Yamaha’s now-eponymous 4-seater revamped golf carts, although as there were no magnetic markers, the tests could not take place autonomously, and were performed with a driver. Which rather defeats the purpose, but hopefully budget can found soon to complete these early tests.

Among this increasing number of tests, ridehailing giant and aspiring autonomous vehicle fleetmaster Uber is having some teething problems setting up shop in Japan. As we saw from early trials in Fukuoka or Toyama Prefecture (covered earlier), Uber’s early experiments in the Japan market have been rebuffed by Japan’s strict shirotaku (or unlicensed cab) laws. Since then Uber Japan has changed tack and teamed up with local Kyoto NPO Ganbaru Furusato Tango-cho that can offer to drive you to your destination if you are not able to do so. As a town with a population 5,400 of which 40% are elderly and where the last taxi companies pulled out in 2008, Tango-cho was slated as a prime location for Uber’s services. The service kicked off May 26th 2016 with 18 cabbies using their own vehicles, with rides available at half the price of normal taxi rides. Uber provides the IT payment and matching backend and in-car tablet/apps, and in return get an unspecified revenue share of the fare.

A year later in May 2017, a report in the Nikkei BP magazine shared results from the experiment. The service was apparently used on average 60 times a month, which translates to about 3-4 rides a month per taxi. Most rides were found to be made in the morning, with 40% of trips outside of Tango-cho to the surrounding areas of Tango City. Cashless trips, of high convenience to city dwellers and business people, was one of the complaints from local citizens used to spending cash when buying local products and services. Whether the experiment will continue and whether it will be automated anytime soon is anyone’s guess, but all the elements are there and the current NPO experiment could be automated along defined routes tomorrow. Or even restricting automation to major thoroughfares that had been previously mapped out with sensors. As long as Uber’s system is able to invest extensively in localization for the Japan market, and kick off automated testing, in the long-term it could very well become one of the go-to automated transportation platforms for Japan’s rural users.