JATO: Tesla, Model 3 Top Global BEV Sales Charts

Tesla was the top-selling EV maker in the world in 2018, followed by China’s BAIC and BYD, renowned data house JATO Dynamics said today in an emailed statement. JATO tracks actual registrations, as opposed to sometimes elastic data published by OEMs.

According to JATO, 230,000 Tesla cars were registered in 2018, followed by 152,000 BAIC BEVs, 95,000 made by BYD, 92,000 million by Nissan, and 64,000 by China’s Zotye. World’s best-selling BEV has been the Tesla Model 3 with 138,000 units registered, followed by 92,000 BAIC EC, and 95,000 Nissan Leaf, JATO says.

It has been occasionally suggested by parties critical to Tesla that many of its new Model 3 could be sitting unsold in lots around the country. JATO’s registration data do not bear that out. Referring to the Model 3, Tesla’s Q4 shareholder letter talked about “nearly 140,000
units sold,” which should be close enough to JATO’s 138,000.

According to JATO, 1.261 million battery electric vehicles were registered 2018 in the 54 countries JATO covers. That's up an amazing 73.5% from the 727,000 BEVs registered in 2017. Tesla's 230,000 units translate into a global BEV market share of 18%. While their sales are impressive, BEVs are far from taking over the world. The global best- seller is the sport utility vehicle. According to JATO data, nearly 30 million SUVs were registered globally in 2018. This gives the SUV a global market share of 36.4%, compared to a global BEV market share of 1.5%

Tesla, Model 3 Top 2018 Battery Electric Vehicle Sales Charts

Tesla was the top-selling EV maker in the world in 2018, followed by China’s BAIC and BYD, renowned data house JATO Dynamics said today in an emailed statement. JATO tracks actual registrations, as opposed to sometimes elastic data published by OEMs.

According to JATO, 230,000 Tesla cars were registered in 2018, followed by 152,000 BAIC BEVs, 95,000 made by BYD, 92,000 million by Nissan, and 94,000 by China’s Zotye. The world’s best-selling BEV model last year the Tesla Model 3 with 138,000 units registered, followed by 92,000 BAIC EC, and 95,000 Nissan Leaf, JATO says.

It has been occasionally suggested by parties critical to Tesla that many of its new Model 3 could be sitting unsold in lots around the country. JATO’s registration data do not bear that out. Referring to the Model 3, Tesla’s Q4 shareholder letter talked about “nearly 140,000 units sold,” which should be close enough to JATO’s 138,000.

Tesla, Model 3 Top 2018 Battery Electric Vehicle Sales Charts

Tesla was the top-selling EV maker in the world in 2018, followed by China’s BAIC and BYD, renowned data house JATO Dynamics said today in an emailed statement. JATO tracks actual registrations, as opposed to sometimes elastic data published by OEMs.

According to JATO, 230,000 Tesla cars were registered in 2018, followed by 152,000 BAIC BEVs, 95,000 made by BYD, 92,000 million by Nissan, and 94,000 by China’s Zotye. The world’s best-selling BEV model last year the Tesla Model 3 with 138,000 units registered, followed by 92,000 BAIC EC, and 95,000 Nissan Leaf, JATO says.

It has been occasionally suggested by parties critical to Tesla that many of its new Model 3 could be sitting unsold in lots around the country. JATO’s registration data do not bear that out. Referring to the Model 3, Tesla’s Q4 shareholder letter talked about “nearly 140,000 units sold,” which should be close enough to JATO’s 138,000.

Monster Map-Making Marriage Gives Birth to Global Standards War

Maps are essential for any strategy. For autonomous cars, maps are the strategy. Away from the headlines, a global war is waged for interpretational jurisdiction in the mapping business: Those who define the standards will reap the rewards. At first glance, a relatively unknown Japanese company bought a little-known American outfit today. Once you look at the bigtime backers of both, you quickly realize that this amounts to a shrewd power grab, and a major re-draw of the mapping space.

Today, Japan’s Dynamic Mapping Platform Co., Ltd (DMP) announced the impending takeover of Canada-born and US-based Ushr Inc. The deal was first announced by Japan’s Nikkei last night, and then followed by a press release coming from Mitsubishi Electric this morning. Why Mitsu? The Japanese electronics and electrical equipment Godzilla is the second-largest shareholder of DMP. The largest shareholder of DMP is INCJ, a consortium of every carmaker in Japan and most of the island nation's big mapmaking and surveying companies. Ushr is backed by none other than General Motors. Ushr’s 3D map data is the secret behind GM’s Super CruiseTM, the industry’s first true hands-free driving technology for the highway, which debuted in the 2018 Cadillac CT6.

Listening to Mitsubishi Electric’s 63-year-old, but eminently youthful CEO Takeshi Sugiyama this afternoon at the company’s HQ in downtown Tokyo, it quickly became clear who is pulling the strings in the marriage. “Both Ushr and Mitsubishi Electric are leaders in high precision maps,” Sugiyama told me, “our systems are a great fit, and the merger will help us establish a global mapping standard.”

Ushr and Mitsubishi Electric have mapping data precise down to a fraction of an inch (10mm) The rest of the industry mostly is proud of maps with a precision of around three feet (one meter). The maps of both Ushr and Mitsubishi Electric far exceed the requirements of an auto industry that thinks 25 centimeters is plenty enough precision.

If anyone is vertically integrated in this business, then it’s Mitsubishi Electric. The company built the satellites for Japan’s Quasi-Zenith Satellite System that is the backbone of the country’s own GPS, it also provides the many earth stations that add augmentation necessary for positioning data down to the centimeter, along with the boxes to ingest the data. Their system excels in the steep urban canyons of Japan’s monster metropolises, where regular GPS often fails. “With our system, the blue dot is where you really are,” grinned Sugiyama. The European partner of the Japanese/American mapping axis is Bosch in Germany, which formed a joint venture with Mitsubishi Electric.

Today’s announcement comes on the heels of another gutsy mapping move. Two weeks ago, Toyota’s high-tech spinoff TRI-AD announced plans for an open source mapping platform, free for everyone on a give and take basis. When I asked Mitsubishi Electric’s spokesfolks how the new mapping-marriage fits in with Toyota’s libertine tendencies, I received no answer, but a big smile.

In somewhat related news, there is hope that the weather no longer will rain on the self-driving parade. While at Mitsubishi Electric, I received a quick glimpse of a system that promises to overcome an all too common problem bedeviling both self-driving cars and old-style analog drivers: Rain and fog can render human vision useless, and it can completely confuse the computer. The elements can even make expensive LIDAR fail. Mitsu-researchers Hiroi Kimihiko and Ryota Sekiguchi combined the inputs of common car sensors to see through pea soup and downpours. Their proof of concept currently runs on a PC-class computer, but the two are confident that they will make the software perform on the less muscular ARM-based systems-on-a-chip favored in the auto business. If your car wants to see clearly now (or rather in a few years when the system will be ready) talk to the Japanese..

The Mystery of Tesla’s Autopilot-less European Model 3

There are recent reports that Tesla’s “Autopilot” isn’t type-approved in Europe, and therefore absent in the first batch fresh off the boat in Belgium. In the EU and in many countries throughout the world, a valid type-approval is essential for the legal operation of a vehicle. The reports of Tesla’s AWOL Autopilot not only caused an uproar on Twitter. The allegedly absent approval also is the source of intensive puzzlement in the auto industry, for one simple reason: There is no EU type approval (yet) covering self-driving technology.

For this article, I could simply draw on my long and occasionally painful experience with the subject matter. It could be outdated, especially when it comes to a both fluid and impenetrable set of EU rules. I therefore called upon a number of homologation experts in the EU, and elsewhere. Not wanting to discuss another OEM, the sources have quested anonymity, and it has been granted.

To a man (and a woman) all experts told me: “There is no type approval for self-driving cars.” Neither in the EU, nor in the U.S., China, or elsewhere. Read the EU regulations, try to sift through all amendments and annexes, and you will find no regulation covering the type-approval of self-driving cars.

You probably begin to wonder how a EU that regulates even the shape of a banana could possibly not have rules for stuff the world has obsessed about for years. Not to worry, it is just a matter of time, I am told. Several working groups, in Europe and at the United Nations, are discussing regulations covering self-driving cars, with results expected in 2020, or more likely 2021. As long as no type-approval for self-driving cars exists, exemptions for self-driving cars will be carved out of traffic laws.

But what about the current Model3?

The experts interviewed were not familiar with the Model 3 type approval, and being experts, they didn’t want to hazard a guess. Some voiced the opinion that regulatory problems with the Model 3 “Autopilot” should be unlikely, simply because Tesla’s Model S and Model X have received EU type approval when equipped with the same Autopilot module.

I don’t give up easily, and taken by the hand by one of the world’s top homologation experts, I traversed the thicket of laws, regulations, amendments and annexes, until my guide pointed to vague references to the technology, buried deep in the United Nation’s Regulation No. 79, covering “the approval of vehicles with regard to steering equipment.” This regulation has been included into EU type approval regulations, and compliance is mandatory.

At first glance, this regulation specifically does not allow “systems, which do not require the presence of a driver, [and which] have been defined as ‘Autonomous Steering Systems’.” No soup for you.

But maybe a cookie:

UN Regulation No. 79 allows “Advanced Driver Assistance Steering Systems” for functions such as Lane Departure Avoidance, lane changes (when initiated by the driver) or systems that “assist the driver in maneuvering the vehicle at low speed in confined spaces.” According to the interpretation of my expert, the current rules allow autonomous steering only at speeds below 10 kilometers per hour, i.e. in an automated parking situation.

Not being familiar with the Model 3, my expert guide refused to give an opinion whether the Model 3’s Autopilot would comply or collide with these regulations.

Tesla’s “Standard Autopilot” comes with automatic emergency braking, along with frontal and side collision warning, and this should easily pass Regulation 79. However, if the Model 3 “Enhanced Autopilot” indeed comes with the features advertised on Tesla’s website, then there is potential of rule breakage. A car that, as promised by Tesla’s website, “takes over, steers itself,” a car that “automatically changes the lane,” a car that “masters freeway exchanges and exits autonomously” could collide with UN regulation 79, and by extension with the EU type approval regime. The final decision would have to be made by the EU regulator, and/or by an accredited technical service, in Tesla’s case RDW in the Netherlands.

Bottom line, Tesla’s Autopilot cannot collide with a type approval for autonomous cars that doesn’t exist in Europe. The Autopilot has, however, the potential of a conflict with provisions deep down in the United Nations regulations on steering.

Russ Mitchell, who covers Tesla for the Loans Angeles Times, cited an RDW spokesman last Thursday as saying that “at this moment the autopilot is not part of the original Type Approval of the Tesla Model 3.” It wasn’t clear which of the two Autopilot versions this referred to. A day later, Mitchell received a sibylline email from the Dutch regulator, stating that “today the approval of Tesla's Model 3 has been published with ‘Autosteer’ added.” Mitchell did not receive amplification, neither from RDW, nor from Tesla. Oddly, those who scour the pertinent United Nation’s Regulation No. 79 will find no mention of “autosteer.” Should “autosteer” refer for Tesla’s lane keeping functionality, then that would be permissible as what the rules call “Automatically commanded steering function (ACSF) Category B1.”

Also on Friday, Bloomberg told its terminal subscribers that Autopilot (which one?) gained “Approval for Autopilot in Europe.” Again oddly, this hot news did not find its way into Bloomberg’s news service and website. At the time of this writing, RDW’s database of type approvals shows no change to the three entries covering Tesla’s Model 3.

Should features of Tesla’s Enhanced Autopilot conflict with the rules, Tesla would be in trouble way beyond Europe. United Nation regulations are observed in countries all around the world, except for the U.S.A. Countries or regions are free to adopt any of the more than 140 UN regulations, and therefore, Regulation 79 is not automatically in effect in every country that subscribes to the UN rules, but it is in many. Should Tesla’s Enhanced Autopilot be held in violation in a Europe with a leading role in the UN regulatory framework, it could have a ripple-effect around the world. Tesla’s solution could be to re-submit a stripped-down, make that nEUtered, version of the Enhanced Autopilot, and it would have to studiously avoid adding the nEUtered features later Over-The-Air. Tesla’s European fans would not be happy.

OTA updates themselves are an even bigger can of worms. As intimated in a previous article, over-the-air updates can conflict with one of the core tenets of UN and EU type approval, namely that a type-approved component may not be altered without a new type approval. As one of the experts told me, “if the software update concerns something that is regulated by Type Approval, an OEM has to get an update to a Type Approval before pushing the software update.” This matter also is in flux, and I was urged to seek the advice of an OTA homologation expert, which I shall do. Stay tuned for more legalese.

Toyota’s Tri-Ad Division Wants Open Source Maps to Guide the Self-Driving Cars of Tomorrow

If Toyota has its way, the maps guiding your future autonomous vehicle won’t be delivered through Google, Apple, Baidu, or TomTom, but via a map platform as free and accessible as Linux, MySQL, Apache, and the zillions of other open source programs powering our lives. That open source map project is called AMP, or in “Autonomous Mapping Platform.”

The concept was presented Thursday by Mandali Khalesi, the man in charge of mapping systems, automated driving strategy, product and alliances of Tri-Ad, the autonomous drive company started by Toyota with Denso and Aisin. Tri-Ad’s mission is to turn the research developed by TRI, or Toyota Research Institute, into products, with the help of nearly 2,000 employees and a $2.8 billion budget.

At Tri-Ad, Khalesi joins an A-team of tech celebrities. CEO James Kuffner was a member of the software engineering team that developed Google’s self-driving car. he company's CTO is Ken Koibuchi, who was in charge of Toyota’s automated driving technologies. The software chief is Nikos Michalakis, who built Netflix’s cloud platform. And UX is in the hands of Thor Lewis, a former Google alumnus.

Like everyone at the top of Tri-Ad, Khalesi is a colorful man. He graduated with a degree in astrophysics; when the light hits him in a certain way, you could mistake him for an alien. He came to Japan in 1999; six years later, he became famous for establishing Japan's first location-based mobile advertising business for luxury brands in a traditional Japanese house in downtown Tokyo. He worked for map companies such as TomTom and Nokia, and headed the Asia-Pacific business of Nokia Here, which later was bought by a consortium of Daimler, Audi and BMW. Toyota then snatched him up as its global head of automated driving mobility and innovation.

Four-dimensional maps are central to automated driving, Khalesi said. “Without a mapping framework, the vehicle is driving blind.” Companies the world over are feverishly amassing mapping data, but then, the data sits mostly tucked away in “sensor silos at OEMs, telematics companies, and mapping companies,” as Khalesi explained.

Depending on where the bulk of drivers drive, the data can get quite tribal. “If I have 100 million miles of test data from California, I won’t have a system that can drive in Mumbai,” Khalesi’s boss James Kuffner said. “Data quantity isn’t the end of story. We need data quality and data diversity.”

Sitting in their silos and accessed through proprietary channels, data cannot enhance and complement other such sets. “What we really need is a one stop open API, something where anybody can access the data from wherever they are by any software or hardware,” Khalesi said.

Unbeknownst to many, Toyota has quietly become a big open source champion. Toyota was the main driver behind Automotive Grade Linux, an open source O/S supported by many OEMs and suppliers around the world. Toyota also is a premium member of the Autoware Foundation, a non-profit supporting open-source projects for self-driving mobility.

Tri-Ad has shown proofs of concept, and in a year or two, you should be able to clone your copy of AMP from Github, Khalesi told me. Once AMP is up and running, Toyota will fill it not only with data generated by the 10 million cars the company sells each year, but eventually with data coming from hundreds of millions of other cars. Anyone is free to use AMP, said Khalesi, “but in return we expect OEMs, fleets, taxi companies, developers to share anonymized data on AMP with other users.” I asked Khalesi where that would leave his former employer Here, which had plans to become the world’s supplier of real-time mapping systems. “They will be welcome to use the system like everybody else,” he said.

Toyota’s initiative to leapfrog the nascent mapping business with a bold open-source project could hit the presumptive mapping giant Here during a delicate phase. There is incessant chatter in Here's new hometown of Berlin that the company's top management could be concentrating on low-hanging—and decidedly non-automotive—fruit, such as credit cards and retail, in an attempt to beautify the company for an IPO. A source close to the matter told me Here's early focus on automotive seems to be getting lost, to the chagrin of the company's backers in the auto industry.

After all, telling cars where to go—autonomous or otherwise—is big business. Khalesi said that map suppliers charge OEMs “a few hundred dollars per car.” There are roughly 100 million automobiles sold worldwide each year. If successful, Tri-As’s AMP could disrupt a $20 billion business before it gets going in earnest.

Swedish Concerns About Tesla’s OTA Updates Could Become Global Nightmare

Twitter is abuzz with the news that Sweden’s NHTSA-equivalent Transportstyrelsen could be looking into issuing a stop sales order for all Tesla cars coming into the Nordic country. What triggered the buzz was a report in Sweden’s reputable Svenska Dagbladet, saying that the regulator is taking issue with Tesla’s over-the-air updating practice. The paper cites the regulator’s spokesman, Anders Gunneriusson as saying: “We have an on-going article-29 procedure on Tesla. This means that we are investigating whether Sweden should stop Tesla from selling cars for six months if there is a serious traffic safety risk.”

This brings a long-simmering controversy into the open: Does Tesla’s over-the-air update (OTA) violate EU rules, and possibly rules affecting large parts of the world, with the exception of the U.S.A.?

Before we get into this, a little background to the Swedish story: The matter is not new, Sweden has been looking into this since 2015. In November of that year, Sweden’s regulator wrote a letter to its partner agency in the Netherlands, RDW (or Rijksdienst voor het Wegverkeer), saying that in its opinion, Tesla’s “Autopilot” should not be approved. A month later, Sweden voiced its concern in a letter to the European Commission, prompting several inconclusive meetings. Now Sweden’s regulator is considering closing the case, but according to a spokesman “if there is an accident in Sweden connected to this, we will intervene and we will open it again.”

This issue is not about Tesla’s autopilot itself colliding with EU rules about autonomous cars. There are no EU rules governing autonomous cars beyond an EU-wide law which famously requires that the driver “must be in control of the vehicle at all times.” This is about Tesla’s OTA updates, and the possibility that they could run afoul of rules governing type approval in the EU and many other countries of this world.

Type approval rules are a thorny matter. To completely transcend them, one must be a combination between a lawyer and an engineer. I am neither, but I received indoctrination when doing communication for a very large German automaker, and I was given a thorough schooling in the subject matter when I was insane enough to enter the automotive parts business for a while.

According to EU law, a car should be tested by the authorities before it is allowed to get on the road. Testing each car would overwhelm even Europe’s massive bureaucracy, and herefore, the regulator is willing to test just one (or a few) cars of the same type, and it issues a type approval as long as the submitting automaker or importer swears that all other cars of the same type will be exactly the same. Once the type approval has been issued, the approved product my not be changed. Of course, technical progress marches on, and engineering changes do happen. Type approval allows them, but only if the type approval receives an approved amendment.

What if Tesla (or any other automaker) sends an update overnight that changes the tested and approved parameters of the car, or adds functionality that wasn’t there when the car received its type approval? According to the letter of the law, that update needs is illegal unless the type approval is amended. The spokesman for Sweden’s regulator reiterated this rule to the Svenska Dagbladet.

According to the Svenska Dagbladet, carmakers like BMW and Mercedes make changes to the software only after receiving approval. “Tesla is not using the same procedure,” the paper writes.

The fact that OTA collides with the letter of the law is nothing new. The matter has been discussed for years (and so far inconclusively) among stakeholders in United Nations working groups, the EU and elsewhere. Carmakers are represented through their global umbrella organization OICA. “Tesla is not a member,” writes the Swedish paper. So why were these concerns never aired publicly? “OEMs don’t want to rock the boat and endanger their own OTA plans,” tells me an informed contact in the type approval business.

The bottom line so far is that Tesla’s OTA is operating in a legal grey zone, and it could get dark rather quickly if the Swedish matter boils over. Should an accident prompt the Swedes to determine that Tesla’s OTA is against the law, “the result would be dramatic,” writes Svenska Dagbladet. As per Europe’s 2007/46 Framework Directive, type approvals issued by a European member state are valid in all of Europe. However, Article 29 of that directive specifies that if there is a serious risk to road safety, any “Member State may, for a maximum period of six months, refuse to register such vehicles or to permit the sale or entry into service in its territory of such vehicles.”

If Sweden would do that, the matter could snowball through Europe and beyond. According to Article 29, the European Commission would need to be involved, and the Commission would have a serious discussion with the agency that issued the type approval, in this case RDW. This could theoretically lead to an EU-wide loss of the type approval. A more likely outcome would be that Tesla is compelled to secure approval for each and every OTA update before it is performed, which also could involve supplying Tesla EU customers with the paperwork documenting the change. In many EU countries, notably in Germany, changes or re-fits “that change the functionality of a part of a vehicle” need to be entered into the car’s registration papers. Should the matter come to a boil, it is quite possible that a nightly OTA update require all Tesla customers to visit their motor vehicles department to update their papers (cost between 11 and 31.50 Euro.)

The issue could even turn into a world-wide problem. Europe’s type approval system is based on United Nations (UNECE) regulations, and so are type approvals of many countries the world over. The rule precluding after-the-fact changes is one of the core tenets of the UNECE agreement with the easy to remember title “Harmonized Technical United Nations Regulations for Wheeled Vehicles, Equipment and Parts which can be Fitted and/or be Used on Wheeled Vehicles and the Conditions for Reciprocal Recognition of Approvals Granted on the Basis of these United Nations Regulations.”

If the genie cannot stay contained in a Swedish bottle, OTA could turn into a worldwide nightmare. Tesla better pray that their cars don’t collide with any elks while in Sweden.

Your Next Car Could Have A Firewall

Want to learn how to hack a car? How-tos teaching the hacking of a car’s internal network are the highlight of any Blackhat confab, and they are readily available on Youtube. A fuzzy video of a Tesla Model X getting pwned in China is always a sure way to make headlines. Dark thoughts of hackers messing with our cars trigger deep seating fears of losing control of our lives. Finally, the empire strikes back. Developments are underway to ringfence our car computers with firewalls and intrusion detection systems like those in big datacenters. Today, I received a first glimpse of them.

Cars have been “hacked” pretty much since the electronic KE-Jetronic obsoleted the mechanical K-Jetronic used to inject gasoline into our engines. When CPUs in our cars started to converse via a network called CAN-bus (as in Controller Area Network) hackers quickly hopped on that bus. Modeled after the infamous Wireshark network snooper, a CAN-bus centric CarShark tool soon followed, allowing hackers to override brakes, pop hoods, or to disable the car completely. 10 years later, the venerable CarShark is pretty much dead, but a huge selection of car hacking hardware and software has taken its place.

However, CAN bus hacking required physical access to the car. Old cars were not part of the Internet. The hacker had to plug something into the OBD (on board diagnostics) port of our cars and sit next to us while putting his and our lives at risk with a laptop. These days, many cars are on-line 24/7, and they are at risk of being broken into with something more insidious than the old Slim Jim. It was only a matter of time for a car’s entertainment system to be infected with a ransomware virus, gladly only as a proof of concept --- so far.

This morning, Dr. Tetsuo Nakakawaji, leader of Mitsubishi Electric’s IT R&D Center, showed me what his company intends to introduce to keep hackers out of our cars. What he showed me looked pretty much like the layered defense surrounding commercial data centers or critical infrastructure: Intrusion Detection Systems (IDS) and firewalls protecting each ring of the system. Many IDS are signature-based, which exposes the system to a new threat until it has been found, documented, and programmed into the system. Mitsubishi’s IDS is heuristic, it looks for “abnormal” activity, and blocks it.

“So?” some will say now, and right they are. None of this is new, and it could be achieved by sticking a Cisco Firepower 9000 (or a pfSense box for the financially less endowed) into our trunks, and by calling it a day.

Not exactly. For price and power consumption reasons, the computers toiling in our cars typically are quite lightweight. A favorite car-CPU is an Arm-based System-on-Chip, something like a Raspberry Pi we can buy for $35. Mitsubishi Electric doesn’t sell boxes, and it wants to run its firewalls and IDS systems on the weak-chested chips favored by OEMs. To get speed, Mitsubishi replaced raw computing power with higher-performance software.

Despite some needling, Nakakawaji wouldn’t say more about the code than that it runs very well on very modest silicon, and a test was not provided this morning.

Another example of Mitsubishi Electric’s in-car coding provided a little glimpse into their more-with-less activities. As we all are painfully aware, computers can take time to boot. Customer don’t want to watch twirly boot animations, customers want the thing to be up and ready when the go-button is pushed. To get around the boot time, developers at carmakers use fast boot tricks akin to the sleep functions on our laptops. An image of the memory is created, and stored, and next time, the image is simply shot into the computer instead of going through a laborious boot process. So far, so good. Or maybe not, because that image could be tampered with while we and the computer are sleeping, and next time we fire up the car, it will be hacked.

The solution is to check the image before it is used. The image is read-in byte-by-byte, and using cryptographic routines beyond the scope of this article, the image is compared with what we had before we sent car and computer to sleep. That, however, takes time, especially on an anemic Arm, and it defeats a good chunk of the exercise’s purpose. Instead of checking the whole boot image byte-by-byte, Mitsubishi’s system reads only choice parts, and it “minimizes read time using a trusted tree method,” as Nakakawaji told me this morning while carefully checking whether my eyes would glaze over. As you can see from the link injected into his quote, trusted tree is not a Mitsubishi invention, and Nakakawaji readily concedes it. “The method is standardized, but our implementation is unique in the field.”

But is it really faster, and how much?

“We haven't run actual benchmarks against the competition, but we definitely hope to be faster,” Nakakawaji said. Well, in Japan, bragging is considered uncouth, and we’ll see it when we see it.

Mitsubishi Creates Smart Navigation That Listens

Does this sound familiar? You are threading your car through a dense neighborhood as your nav system announces:

“Left turn in 300 meters” (or yards, leagues, furlongs, whatever.)

You peer through the windshield, and mutter: “This one? Or the next one?”

The nav system can’t hear you, and it leaves you guessing.

The matter is especially pressing in a country like Japan, with its warren of little streets and similar-looking houses. So it fell to a Japanese company to finally come up with a solution.

As a preview of a broader announcement next month, the solution was shown to me this morning in a downtown Tokyo meeting room of Mitsubishi Electric

Again, the driver muttered: “This one? Or the next one?”

Except this time, a human-sounding, but still artificial voice answered: “Not this one, the next one.”

The immensely helpful system is quite tricky to implement properly, Mitsubishi engineers explained to me. “Cameras monitor the lips and eyes of the driver, to see who is talking and where the person looks,” told me Masahiro Abukawa, General Manager of Mitsubishi’s Human Intelligence Technology Department.

Detection of who’s talking is necessary to filter out comments by backseat drivers not worthy of computerized assistance. When kids in the rear asked, “are we there yet?” Mitsubishi’s voice robot wisely remains silent. To help the robotic brain understand what the muttered “This one? Or the next one?” refers to, it detects where the driver is looking.

Also demoed was another urban jungle survival tool: little cameras mounted into the nose of the car (or repurposed sensors from the existing suite) that can tell the driver if the road is clear when edging out of a garage or narrow alleyway. Cameras and corresponding small monitors are engaged via the driver’s eye movement.

Both technologies are building blocks in Mitsubishi Electric’s MAISART compact artificial-intelligence technology for smart mobility, the company says. Rather than trying to do away with the driver, MAISART, like other ADAS technologies by Japanese OEMs like Toyota or Nissan, is focused on assisting the driver as much as possible, and on turning her or him into a bionic driver with capabilities unsurpassed by man or woman. This approach is essentially the "augmented driving" concept championed by The Drive's Alex Roy.

You can buy the MAISART stack if you are an OEM, or a component manufacturer. Mitsubishi licenses the technology instead of putting it into own hardware.

Exclusive: Toyota Hydrogen Boss Explains How Fuel Cells Can Achieve Corolla Costs

If you want to bring the wrath of Twitter upon you, suggest that electric vehicles could be powered by something other than batteries. Only the utterly naïve seem to think that any technology bringing zero emission transportation should be worth the consideration. Mention fuel cells, and you will swiftly be denounced as an idiot, an agent of Big Oil, or worse. Tesla’s Elon Musk routinely piles invective on fuel cells, calling them "mind-bogglingly stupid," "incredibly dumb" and "fool cells”—schoolyard taunts that resonate especially well with battery-bullies.

The other side of the holy war for the most part quietly turns the other cheek. Well, until this morning, as I sit down in a coffee shop in Tokyo’s Shinagawa Station for an exclusive interview with one of the central figures of Toyota’s fuel cell project. Professor Katsuhiko Hirose has been in charge of Toyota’s Fuel Cell system development, and he tells me why Musk is rightfully afraid of the fuel cell: Hydrogen Fuel cell vehicles will eventually be cheaper than gasoline-powered cars, opening the door to zero-emission transportation around the world, and not just in rich pockets that can afford battery-electric vehicles that are dependent on scarce natural resources and government support.

Hydrogen fuel stations are few and far between, and this scarcity opens the door for skepticism of the future of fuel cell cars. Hirose was a young physicist who had traded nuclear fusion for an engineering job at Toyota, where he developed early digital engine controller units. This turned into an assignment of managing fuel economy and emissions for a pie-in-the-sky project that became the hybrid Toyota Prius. When a skunkworks team at Toyota started looking into hydrogen fuel cells, Hirose joined the ranks of people pooh-pooing the project. “I am a scientist, and I pointed out that hydrogen was an energy that needed energy to be produced. I questioned where the hydrogen would come from, and where it would be stored. There were no proper and safe large-scale storage systems at the time,” Hirose remembered. “Ironically, I later I became the person to develop advanced hydrogen storage.”

That came when in 2003 Hirose’s boss Mitsuhisa Kato told him that “it’s time to join the enemy” and switch from hybrid to working on hydrogen. Hirose was made General Manager of Toyota’s fuel cell systems development, and he had to work on dispelling his own doubts. He wasn’t the only doubter.

Three years later, in 2006, the German scientist Ulf Bossel wrote a paper that basically claimed that that making hydrogen from electricity is a waste of energy. Storing the same quantity of energy in a battery rather than turning it into hydrogen for storage and later use is more efficient, Bossel posited. Decades later, the battery cam still treats this paper as if it was brought down from Mount Sinai on two iPads. To this day, Bossel’s screed remains the ideological basis of pure battery doctrine.

Bossel made a fundamental mistake in his analysis, Hirose says about the fellow nuclear physicist. Bossel assumed Hydrogen would be made using existing electricity. “If the starting point is that the electricity is already there, and if you use that existing electricity to make hydrogen with electrolysis, then yes, you would lose 30% of that energy,” Hirose says. “But people forget that when that electricity was made from natural gas, or coal, 60% to 70% of the energy was lost. If hydrogen is produced directly from natural gas or even coal, they will not lose 60 or 70% of the energy. It all depends on where you start.”

The bulk of today’s hydrogen is not made from electrolysis, but by reforming mostly natural gas, or even coal. Strangely, Bossel did not, as he himself admitted, “consider this option, although hydrogen can be chemically synthesized at relatively low cost.”

Another mainstay of the counter-fuel-cell dogma is that "when storing liquid hydrogen, some gas must be allowed to evaporate for safety reasons, meaning that after two weeks, a car would lose half of its fuel, even when not being driven.” No wonder that hydrogen loses out if we let half of it evaporate before it can power a vehicle. Ah, but Bossel is talking about “liquid” hydrogen, and Hirose calls him on it.

“Automakers have long given up the idea of using liquid hydrogen in their passenger cars,” exactly because their small tanks lack sufficient insulation for the minus 253 Centigrades of the liquid hydrogen, and half of it could indeed boil away within two weeks, I am told. The hydrogen used in fuel cell cars is pressurized, not liquid, and it won’t escape. However, it may escape those who still copypaste a 13-year-old, and long outdated paper.

Large scale storage of liquid hydrogen, on the other hand, is extremely practical, I learn over a second coffee from Toyota’s expert for advanced hydrogen storage systems: “At Japan’s space station in Tanegashima, enough liquid hydrogen to power a number of rocket launches is stored for many months with evaporation of less than 1 % of volume.” Even that is not wasted, and is used for power generation or other applications.

Another matter escaping the author of the alleged analysis was the heft of the battery. Weight is the enemy of mileage, and trying to gain range by adding more battery quickly becomes a losing proposition as more and more energy is wasted dragging a heavy battery around. Nowhere does this hurt more than with the electric trucks planned by several automakers.

“A battery-electric 40 ton truck with 500km range needs 8 tons of battery. That’s ridiculous,” says Hirose. “You want to transport goods, not a huge battery. A fuel cell stack is much lighter and easier to handle.” A hydrogen fuel cell powertrain’s weight is comparable to that of a diesel engine, Hirose tells me, sometimes it’s even lighter. “Per weight, hydrogen delivers three times the energy of diesel fuel, but it also needs a heavier tank,” Hirose says.

Times and technologies have changed since Ulf Bossel wrote the paper, and he himself has long departed from the pure battery faith. Ulf Bossel is now owner of the Swiss Almus AG, which sells the UBOCELL, a small SOFC fuel cell that turns hydrogen into electricity.

The biggest threat to the battery camp has been kept under cover so far – until this morning when I order my third coffee: Fuel cell vehicles will soon be cheaper than Toyota’s own hybrid cars, and they may one day become cheaper than regular cars. That may come as a jaw-dropper, after all, haven’t we been told that fuel cell cars are inordinately expensive, because they use high amounts of platinum?

Nonsense, says Hirose: “Back when the first fuel cell vehicles supposedly cost a million dollars, we used 100 gram of Platinum in them. You can buy that for $3,000. Now we are using …” And he pauses. “Much, much less.” A back and forth ensues on how little Platinum goes into Toyota’s fuel cell stack, and finally, Hirose reveals that it is in the neighborhood of the platinum in a catalytic converter for diesel cars, which “uses around 10 grams of Platinum.”

“People still think fuel cells use a lot of expensive precious metals. Not true,” says Hirose. “The most expensive stuff of the fuel cell stack is a 0.01 millimeter thin membrane. The rest of the stack is very cheap. Actually, the most expensive part of the fuel cell is the process, it’s the making of the stack, not the materials going into it.”

That puts the fuel cell at a huge cost advantage vis-a-vis the battery. “70% of the cost of the battery is in the raw materials,” says Hirose, and the price of raw materials usually does not come down as demand goes up. Neither do batteries free us from being tied to political hotbeds. A key ingredient of batteries is Cobalt, and 66% of the world supply comes from the allegedly Democratic Republic of Congo, where it is mined amid continuous complaints about human rights abuses. Cobalt prices have been exploding, and OEMs are in a rush to secure supplies. Lithium is plentiful in comparison, but nonetheless, The Telegraph reportsof “a frenzy of activity as a global scramble erupts to extract the metal and secure supplies for lithium-ion batteries.” As long as there is water (H2O, two atoms of Hydrogen for one of Oxygen) in the world, there is no way to run out of Hydrogen.

With 70% of the battery cost tied to raw materials, cost optimizations must focus on the remaining 30%, and “the battery is already scaled out,” says Hirose. But haven’t battery costs come down a lot? “Battery cost has come down drastically because of overcapacity,” Hirose says. “Batteries are a commodity, and if someone sells them 10% cheaper, they get the sale, increasingly below cost.” The huge battery factories built around the world require monstrous up-front investments, and advances in technology leave little time to recoup the CAPEX. Toyota’s battery partner Panasonic wrote “$7 bln in losses in two years, mainly over the write-down of an outdated battery factory,” Hirose says, “and many battery makers are in the same situation.”

Panasonic makes the battery cells that Tesla then turns into battery modules in its Nevada Gigafactory. Panasonic is “still losing money” with Tesla, Panasonic President Kazuhiro Tsuga recently told Bloomberg, saying he was hopeful it would change. Meanwhile, the round 18650 and 2170 cells Panasonic makes for Tesla are already outdated, and when Toyota and Panasonic announced an alliance to first make prismatic cells, and then solid state, Tsuga was publicly hoping that technology would not advance too quickly, saying that “If we would have shift to solid state batteries all in a sudden, our investments would be wasted.”

In comparison to batteries, “fuel cells still are in the very early stages of scale,” says Hirose. Toyota makes around 3,000 of its $60,000 Mirai fuel cell car, but it is “building a new factory to increase the production to 30,000 per year by 2020,” Hirose tells me. On top of that, Toyota supplies its fuel cell technology to trucks, buses, and “many applications beyond mobility,” Hirose says, a scale that already allows Toyota to have “the lowest cost fuel cell in the world.”

It is Toyota’s plan to bring the price of fuel cell cars to the level of hybrid cars by 2025, but “personally, I am not very comfortable with that target,” Hirose admits. “To change the world into alternative fuel, we need to provide the fuel cell cheaper than conventional vehicles. That is my personal objective. I think it can be done.”

This may sound like an outrageous claim, but according to fuel cell engineer Matthew Klippenstein, who is also co-author of the industry-standard Fuel Cell Industry Review, "Toyota's optimism on fuel cell costs is quietly shared in the industry, even as batteries have deservedly earned headlines. As the production of fuel cells scales, their cost will plummet.”

I ask Hirose when he thinks that a fuel cell car will cost less than a Corolla.

After a pause, he says:

“It depends on how serious we are about truly sustainable transportation. It can’t be done only by Toyota. If society is willing to provide the resources, it’s possible. I’m now 63 years old, I think I will live to see it.”

The debate about the fine nuances between batteries and fuel cell stacks has not ended, and it will rage on as long as factories have to be amortized, and as long as there is government money no one wants to share with the other camp. While battery power gets all the headlines, support for hydrogen is growing around the world. During the World Economic Forum 2017 in Davos, a Hydrogen Council was founded with 53 energy, transport and industry companies as members, and with Hirose as Toyota’s spokesperson in that forum. China announced plans for 350 hydrogen stations by 2025. Germany wants to have 400 stations by that date. Japan threw the heft of the world’s third largest economy behind fuel cells and announced plans to turn the island nation into a “hydrogen society” anywhere between 2030 and 2050. Japan’s capital Tokyo wants hydrogen in a greater hurry, and aims to realize the hydrogen society by 2020, when the Olympics roll into town next year.

Hirose is an outspoken man, and he doesn’t “like the term hydrogen society,” as he admits. “Hydrogen society means we fully bet on hydrogen. Instead, we should bet on a portfolio of solutions for a sustainable society. Of course, the government can incentivize, but in the long term, we need to provide what the customer likes to choose. This is not about batteries vs. fuel cell. It is about sustainable, zero emission transportation, and about customer choice.”

“A sustainable society is a lot of work,” Hirose says as he collects his things to catch his Shinkansen back to Nagoya, “but if we postpone decisions about matters like global warming, we steal the future of our children. As a company man, I should not say this,” he tells me, “but maybe you can.”

I just did.