Nissan, DeNA expect early 2020s self-driving taxi rollout
An Easy Ride robotaxi waits at a pickup point. Photo credit: NAOTO OKAMURA
TOKYO — Nissan Motor Co. is deepening collaboration with a Japanese mobile gaming and communications giant to develop self-driving taxis.
Nissan wants the partnership to lift it ahead of its rivals in the nascent vehicle mobility services segment. But when it comes to realizing self-driving taxis that can pick up and drop off customers automatically on public roads without a glitch, the automaker acknowledges there is still a long way to go.
Last month, Nissan Motor Co. and online tech company DeNA Co. field-tested Easy Ride robotaxis, involving some 300 participants.
“This represents a big step toward enhancing self-driving cars and the mobility service operation system from the stage of just presenting a conceptual image,” Nissan CEO Hiroto Saikawa said during the launch. “This will help advance our business in offering a new mobility service for many customers in a variety of scenes.”
The robotaxis are based on a modified Nissan Leaf electric vehicle. For their field testing, they traveled about two miles on a preset city route from Nissan’s headquarters in Yokohama to a nearby commercial facility.
Using a DeNA-designed smartphone app, users could hail a taxi by selecting a pickup time slot and specifying which of a list of preset destinations they wanted. A tablet computer installed inside the vehicle notified the passenger about recommended events in the area.
Discount coupons for restaurants affiliated with Easy Ride were sent to the user’s smartphone.
Such an online user experience could not have been possible without the help of DeNA, Nissan officials said.
“Efficient and effective collaboration with partners who have expertise in their respective fields are the key to remaining competitive in the future,” Saikawa said. DeNA CEO Isao Moriyasu stressed that his company wants to bring innovation to the transportation system as a mobility service provider.
The companies plan to introduce their robotaxi service commercially in the early 2020s. But they must resolve technical details before introducing a fully autonomous mobility service in a heavily congested urban environment.
Easy Ride uses Seamless Autonomous Mobility, developed by Nissan from NASA technology, for its fleet operation system. That system allows vehicles to make decisions in unpredictable situations with the combined support of in-vehicle artificial intelligence and humans, according to Nissan. There will be a control center where people monitor Easy Ride fleets.
But technology challenges remain. Among them: perfecting the robotaxi’s ability to judge where it is most appropriate for loading and unloading passengers. Unexpected complications, such as road construction or a line of parked cars on the roadside, for example, wouldn’t faze a human driver. But altering behavior to accommodate unusual circumstances doesn’t come easily to an AI-powered, sensor-directed taxi without the aid of remote monitoring by humans at the control center.
At a media test ride in mid-February, an Easy Ride vehicle unexpectedly stopped its self-driving mode just as it was about to get moving when a pedestrian walked in front of it. As a result, another test vehicle was brought in and the procedures had to be repeated.
Nissan expected such glitches to occur during the field test, and is looking to incorporate the experiences into its development.
Kazuhiro Doi, a global director of Nissan’s research division, admitted that the experiment was challenging, even in Yokohama’s waterfront area, with its wide streets and relatively light traffic.
“Self-driving while trying to avoid so many parked cars is actually difficult,” Doi said of the research. “GPS signals are weak or cut off due to high-rise buildings. I thought it would be easy at first, but it has proven to be more challenging.”