Koji Yamamoto of Nissan explores the challenges facing increased autonomy with Masahide Tomonari [Mob.Tomonari.2016.03.29]
As autonomous driving and sharing services advance worldwide, the land of mobility finds large changes arriving at its border. How do industry-leading companies respond to these looming changes?
TU-Automotive Japansat down in conversation with Koji Yamamoto, division general manager of the vehicle-IT and autonomous drive division at Nissan to learn more about the focus areas and strengths at his company and preparations being made in response to change.
Q: What current trends are you focused on?
“We pay attention to the various businesses seizing the era of autonomous driving and EVs, and attempting entry into the realm of mobility, even though these companies are not automotive OEMs. Particularly, the business model of car sharing and autonomous driving without the purchase of an automobile, like Google, may become a threat to the OEMs in the future. Nonetheless, as OEMs we have the record and expertise of manufacturing cars and are developing electric vehicles and autonomous driving. As long as the OEMs steadfastly accumulate technology, they should be able to ride the waves and maintain competitiveness. It’s important to conduct thorough research and unfailingly react to the moment when technology breakthroughs occur.”
Q: Piloted Drive 1.0 launches this year so what’s special about it?
“As its biggest benefit, autonomous driving brings safe, comfortable motoring to everyone: to our motorist customers and to those who are less adept at driving. Stage by stage, or scene by scene as we call it, the enabled range of autonomous driving will expand. Piloted Drive 1.0 will enable automated braking and steering within a single lane of an expressway. The succeeding Piloted Drive 2.0 will have a complete 360-degree view of the vehicle’s periphery. If conditions are safe along a multi-lane expressway, lane changes and conducting the vehicle according to navigation instructions from point-to-point will be accomplished. Today, the driver is required to be hands-on for steering. If legislation is amended around 2018 or 2019 to enable hands-off, then Piloted Drive 2.0 can make an extremely large contribution.
“Then the world opens up for elderly people, handicapped people, and further to people who hold no driving license, and provides benefits. We’re talking about a timeline up until 2020 but the future holds a new world even further ahead.”
Q: What is the current validation testing focused on?
“Various validation efforts are underway with an eye on Piloted Drive 2.0 and 3.0; however, validation and verification efforts are particularly emphasising the market release of Piloted Drive 1.0. For instance, a variety of scenes are employed to verify the on-screen appearance of the dashboard. We are concerned whether the display truly provides ease of comprehension to the customers and whether autonomous driving is understood. In this context, we have people other than professional drivers engaged in the verification. Hence, safe and easy comprehension of the user interface is extremely important.”
Q: Where do you think the technological challenges lie in the next evolution of autonomous driving?
“The depth of technology and direction of development vary according to the stage or scene for autonomous driving and delegation level (how much autonomous driving is assigned to the car). The range of back-up and fail-safe functionality will increase for autonomous driving.
“Further along, the level where the motorist customer is riding in the vehicle and reading emails, for example, will require the vehicle to completely secure safety. We call this ‘eyes off’, which means Level 4 autonomous driving. That level calls for the OEM to warrant safety, and presents extremely high hurdles. Our technological challenges, therefore, vary according to the differences in delegation level and scene.
“Accumulation of empirical knowledge and automated learning become important, for example, through AI that can learn the appropriate lane-change timings. Nonetheless, limitations abound. Discrete solutions and control methods to overcome over 1,000 scenes must be thoroughly thought through. Some scenes will most likely be extremely difficult, so control will have to be returned to the driver.
“The foregoing does not mean that many sensors become necessary. In fact, how quickly validation with AI can be done and enable machine learning is a necessary part. Like the smartphone, OTA upgradability for firmware will be mandatory technology.”
Q: How do you foresee the social issues?
“In terms of legislation, Japan and Europe do not permit hands-off driving and such behaviour will result in an infraction for violating safe driving. The laws need to change for autonomous driving but that means international laws like amending the Geneva Convention and the Vienna Convention first. The same applies to “eyes off” as well. Driverless vehicles are impossible today, however, acceptance may advance with legislation.
“The other side of the equation is social acceptance. How do we change the mind-set of customers who think that autonomous driving is dangerous? An important response is to release autonomous driving soon and get customers to try it. By learning how safe and secure it actually is, I think we can lay the groundwork for acceptance.
“In terms of infrastructure, our first request is to have the lines clearly painted on the road, so that autonomous driving sensors need not contend with blind sections. For safer cruising, however, we may need to discuss the establishment of dedicated autonomous drive lanes. Automatic deceleration is another device needed whenever an accident has occurred ahead in the direction of travel. OEMs would also find different laws and infrastructure among various nations burdensome. Even V2X communication protocols should be implemented with global standards.”
Q: Will OEMs turn into service providers?
“If such an era were to arrive, we must consider a decline in total industrial volume (TIV) for vehicles, since customers who share instead of purchasing a vehicle will occupy a significant share. Probably not half but TIV declining would cause some companies to become unprofitable under solely manufacturing. If the cars built were extremely advanced, however, that part alone might be a solid business, sort of like the iPhone is today.
“Those companies who cannot secure profitability by manufacturing alone must naturally pivot their business toward a service scenario. If you were to ask whether Nissan was contemplating that move, I would say sure, there’s a tremendous opportunity there but we don’t have a strong resolve to migrate.”
18 Oct 2016 - 19 Oct 2016, Tokyo, Japan
Having been dedicated to connecting key figures across the telematics ecosystem in Japan and beyond, TU-Automotive Japan 2016 will explore issues, opportunities and technologies in the connected car, autonomous vehicles and auto mobility.