Osaka: Google, Toyota and 2025

Earlier this year, with a longstanding invitation from our friends newly moved to Osaka, we decided with my wife (and 2-month-old baby!) to catch the shinkansen for a weekend with our friends. Thinking back, last time I was in Osaka was 2001 about to meet my mother at the downtown youth hostel for her very first trip to Japan, prior to accompanying her on our travels to the west and then south of Japan. I was surprised then as now with the sheer amount of construction ongoing, the second largest conurbation in Japan and a city in constant reinvention of itself. So with that in mind, when researching Osaka it was unsurprising to hear that the city has worked or is working with two giants in the automated driving race at polar opposites. On one side of the ring, established leader Toyota with its focus on safety, inventor of just in time, and the Lexus phenomenon. On the other, leading IT industry challenger Google who effectively dominate the search industry and then upended the auto industry with its focus on full-automation and launch of iconic pod-like Google cars back in 2015.

Starting in January 2014, the Yomiuri Osaka newspaper reported a tie-up being worked on between the City of Osaka and Google to test its vehicles on the three islands of Sakishima, Yumeshima and Maishima in the Osaka Bay area. Similar to the current developments around Tokyo’s Haneda Airport we wrote about previously, the City was planning to create innovation zones on the islands for robotics, automated driving and other new technologies. The next step was to get the police and other governmental agencies involved. At the time, Google was still developing automated vehicles using the Lexus platform kitted out with Lidars, cameras and other sensors, and preceded the launch of the iconic Firefly no-handle no-brakes self-driving model by about a year. But after all the excitement, nothing happened.

Was the planned shift from experimental Lexus plus test driver in 2014 to fully-automated Firefly with no steering wheel in 2015 too much for the local/national authorities? Note that while the police has put in place regulation allowing for fully automated vehicle testing in Japan in the last few months, at the time there was no framework for such tests and even in mid 2016 a statement from the Japanese Police reconfirmed no plans to allow such vehicles on the road. Or could it be that when Google decided to spinout its research division (into what is now Waymo), all collaborations were revised globally? Or was it triggered by the departure of the previous director and automated driving guru Chris Urmson? Whatever the reason behind the silence, there have been no updates ever since. Even the original Yomiuri newspaper article sharing details of the announcement has been taken down, showing a barebones 404 page instead.

Fast forward to late 2016 when electronics giant and automotive Tier-1 Panasonic jumped into the fray. Considering their ongoing tie-up with automated driving innovator Tesla, one would have imagined it would be presenting a number of experiments or vehicle platforms. Yet it was reported to be planning the launch of micro mobility-type vehicle tests both in the US and near their headquarters in Osaka’s Kadoma City. In order not to compete with the carmakers at their own game, Panasonic vehicle use cases would be limited to 1-2km routes, ie local trips to the supermarket and other shops. But Toyota’s micro mobility vehicles, such as their COMS vehicles, play this role very nicely today as I found out when I travelled across Ishinomaki City in Tohoku earlier this year. As locals already use their own “kei” or “light” vehicles for transportation, micro vehicles such as COMs are of most interest to local city officials. The use case is to be able to travel 10-20km at short notice from their local town hall office to meet citizens or accomplish tasks that don’t require a full-size vehicle, but is too tiring or too far to travel on foot or by bicycle. As late as November 2016, Panasonic was reported testing its micro EV platform with plans for launch in 2020. However in the July 2017 edition of Japan’s Automotive Research Institute yearly report on ITS just out a few days ago, Panasonic was quoted that they wouldn’t work on building vehicles themselves, rather focusing on the systems for those vehicles. And that testing would take place within their compound and not on the main roads. As we move towards an EV-dominated future with increasing rationalization of body components, it is very likely that 3d printing the vehicle body will become easier and easier. And it is equally likely that the systems and software in the vehicle will win the bulk of the automotive value-add business in the near future. This strategy reduces the company’s future R&D cost profile, keeps a physical testing platform in-house to try out its systems in the real world, and stays focused on its core business – advanced electronics systems. A smart move on Panasonic’s part.

In the background, Toyota has also been busy. As part of the Osaka City ICT Strategy kicked off in August 2014, Toyota has teamed up with the city to offer big data services covering driving and probe data from its vehicles as well as participating partners (although it is unclear whether this means Toyota-affiliated companies Toyota holds stock in, or independent software/data suppliers such as taxi companies). The idea was to offer vehicle sensor-generated information to the city to create a hazard map that could eventually be used by others. Over two years from 2015-2016, Toyota’s IT Development Center collected hard braking information, hard steering events, speed, acceleration as well as current location from its own vehicles. One of the partners was named as Yamato, one of the largest courier services in Japan, whose drivers also provided information on hazardous areas to drive. This resulted in the creation of a “hiarihat” map, which can offer a predictive hazard map to citizens as well as vehicles. Looking at the link, I can count 40-odd locations in the centre of Osaka in the vicinity of the main train station and adjacent areas, with locations ranging from level 1 to level 3 in danger level and a short explanation. However it is surprising that only 40 or so locations made it to the final list after 2 years of experimentation. Surely there is more? Or maybe these locations change on a daily basis? While interesting for a demo, it would be great if this kind of map was scaled nationally and be updated 24×7.

Building on these experiments, while Tokyo has the 2020 Olympics to prepare for, Osaka is setting its sights on the next big event – the World Exhibition of 2025. Currently up against Paris in the early stages of its bid, the idea is to revive Yumeshima in the Bay area as the site for the exhibition. Focused on next-gen healthcare, AI, drones, automated driving and other futuristic technologies, according to the organizers the event would bring 28-30m visitors to Japan and provide a $17bn boost to the national economy for an outlay of only $1.8bn – an impressive 9x return on investment. And while plans are still at the early stage, there is nothing stopping the local government designating Yumeshima tomorrow as a regulatory sandbox for automated driving, as in the original concept discussed with Google back in 2014. By showing concrete progress on real-world testing in the Bay Area, Osaka would not only present a more convincing case to win the Expo but also start building an ecosystem of startups, universities and established players all focused on building cool and innovative technologies. Surely that will be the main draw for 2025? Currently French President Macron is trying his best to attract talent to his country, making the most of the Brexit confusion and making it easier for enterprises and startups to invest and relocate to France. Can Osaka up the stakes and beat Paris? Game on!