Mobileye moves on to mapping, driving policy

Prof. Amnon Shashua (center) at CES Photo: PR

Mobileye chairman Prof. Amnon Shashua on the latest breakthroughs that make his company a unique partner in driverless car development.

It was cold last Wednesday afternoon in Las Vegas. The dozens waiting in line for a spin in the autonomous car developed by Israeli company Mobileye(NYSE: MBLY) and US company Delphi Automotive, however, were indifferent to the dry desert cold. Three beautifully polished Audi SQ5s were parked near the Delphi pavilion in the CES exhibition, within walking distance of the Las Vegas Strip. Delphi representatives watched the cars with a hawk's eye, as if one of those waiting might, in a fit of daring, sneak into one of the cars without attracting attention and drive off for a wild autonomous trip in the Nevada desert. We took part in one of those test runs, squeezing into a rear seat behind Anthony, the driver. "Anthony is our driver today, but he doesn't drive; he just supervises," explains the Delphi representative in the seat next to him. Yes, the car is autonomous, but not completely - Grade 4, which requires a driver's presence, not Grade 5, which means full autonomy without a driver being present. "Anthony will take over if he sees a hazard to which the autonomous system is unable to respond with the necessary effectiveness," the Delphi representative explains. "This car is still conservative in the way it thinks. It doesn't take chances."

Anthony's hands are not on the wheel, his feet are not on the pedals, and the car is traveling on the Las Vegas roads. Its sudden braking, mainly when another car enters the same lane, is a little abrupt and not exactly smooth, but the Delphi representative claims that this is true even when the car is not in autonomous mode and Anthony is the one applying the brakes. "Braking like this is always unexpected," she says. Someone not in the driver's seat will not really feel the difference. It's a little like riding in a taxi without a cab driver gabbling away. Still, even in a rear seat, it is hard not to get excited about the fact that the car is driving by itself without the touch of a human hand.

Anyone with even a slightly provincial attitude will find it hard to ignore the fact that some of the brain behind this autonomous car belongs to an Israeli, even a Jerusalem, company named Mobileye. The auto manufacturer is a German company that has existed for 106 years already (Audi, but that is not the story in this case), and Mobileye's partner, Delphi, is a 22 year-old US supplier of diverse auto parts. Mobileye, which emerged from its development center in the Har Hahotzvim industrial zone, now has a $9.1 billion market cap two and a half years after its IPO in the US capital market. As you will see, this offering is very significant to Mobileye cofounder and chairman Prof. Amnon Shashua, the technological brain behind the company - even more so than the occasion of the test drive near the Delphi pavilion, where he hurried to say hello to all those taking part in it.

Mobileye, selected by "Globes" last September as the company of the decade, is a global leader in the development of computer vision and artificial wisdom, data analysis, and location and mapping for the purpose of assisting drivers and autonomous driving. The company's vision is for its technology "to preserve the safety of the travelers on the roads, reduce the risk of accidents, save lives, and bring about a revolution in the driving experience," according to Mobileye.

Mobileye is not the only one. A recent comprehensive report by the Roland Berger research firm highlights Israel's dominance in the global smart car industry - a dominance that cannot be taken for granted, given that unlike Germany, France, and Japan, Israel has no auto production lines. "I understand what you mean," Shashua tells us when we ask for his opinion about this in a special "Globes" interview. "When you think of Israel and the auto industry, you think about the Susita, and that's all. It's not obvious how we became a power," he says with a smile.

"Globes": So how do you explain the report's conclusions?

Shashua: "The auto industry is undergoing a revolution - a switch from a focus on mechanics, metals, and engines to a focus on digitization. In digitization, what distinguishes a good car from a bad one is the number of sensors and the type of software it has. The driving experience is becoming electronics and software-based, and software is something Israel excels in. That's why the study's conclusions don't surprise me."

Still, in order to develop software for cars, don't you have to understand auto production as well as Germany understands it, for example?

"Not necessarily, and Germany isn't a software powerhouse. When you look where software in in the corporate DNA, you find it in the US in Silicon Valley and New York and in Israel. That's it. India and China have a lot of programmers, but they aren't creating innovative software, meaning that they aren't driven by innovation. They're imitators. Europe has a traditional industry of production and engineering, but not software - look at Nokia's downfall. So it doesn't surprise me that the change in the auto industry, in which software is becoming the focus, is making it land in Israel."

Maybe it is because of you. Did you blaze the trail for many other companies?

"The media is giving us this credit. Without Mobileye, General Motors wouldn't have established a development center in Israel. Mobileye is giving a big push to all the startups being founded in the smart car industry, because they have realized that it's possible. After Mobileye showed them the way by becoming the leading company in this field with only hundreds of employees, instead of tens of thousands, and I stood on the same platform as CEOs of Intel and BMW (Shashua is referring to the three companies' joint press conference, T.T.), it comes as no surprise that Israel is becoming dominant in the smart car production chain."

You mentioned startups. Mobileye is doing everything by itself, without acquiring technologies from other companies.

"Yes, we're doing everything by ourselves. After the IPO, we thought it was an opportunity to acquire companies, even in a shares deal, because the company's high value enabled us to buy shares without even feeling it. We looked around and found nothing worthwhile that we couldn't do by ourselves. We've made a quantum leap since then - a real evolution. In the summer of 2014, when we held the IPO, Mobileye was a company with a solution to a product, and the product was a camera with a sensory chip whose only function was to prevent accidents. We were like a Tier 2 supplier then. That was our positioning, where we were, and still are, the best. Now, on the other hand, following the cooperation with BMW we announced last July, Mobileye is no longer just a supplier. We were invited to the BMW conference as a partner, not a supplier, and that's a significant breakthrough. After all, BMW has hundreds of suppliers, but there weren't hundreds of suppliers with it on the speakers' platform. In other words, Mobilieye has gone from being a supplier of a product to being a partner. You could also say that Mobileye has switched from products for preventing accidents to autonomous vehicle technology. That's an industry already, not a product."

Shashua mentions that the company presented road experience management (REM) technology at the previous CES exhibition. This is a mapping technology that makes it possible to obtain data in real time, based on crowdsourcing. "This is a means of creating a high-resolution map, which is an essential precondition for effective low-cost autonomous driving. Crowdsourcing is the only logical way of devising such a map," Shashua says. According to him, completion of the first REM map covering the high-speed roads in every country is slated for late 2018, in cooperation with an auto manufacturer and a map producer. Just before the most recent CES exhibition, Mobileye announced strategic cooperation with Dutch company Here to facilitate crowdsourcing for the creation of high-resolution maps to be used in autonomous driving.

"Mobileye is working with 27 auto manufacturers, and all the data from the cars with Mobileye chips is coming to us for the purpose of creating the maps," Shashua says.

Will this mapping technology be a revenue source in itself?

"Yes, certainly."

Mobileye's aim in this context is to turn the dynamic map into a revenue model separate from the chip or array of chips that it is selling. Every vehicle in which at least one Mobileye camera has been installed will in effect become a company agent (the crowdsourcing principle), and will help the company map the world dynamically, in other words, to build a very responsive map. Licenses to use this map will be sold to anyone who wants to install an autonomous driving solution.

At the recent CES exhibition, Mobileye achieved another breakthrough with its emphasis on the "intelligence behind merging into traffic," according to Shashua. "This is what we call a 'driving policy' - the third layer of autonomous driving after sensing and mapping."


"When you drive in dense traffic, you're negotiating, not with words or eye movement, but with vehicle movement. Your movement is what signals your intentions to someone to whom you want to yield the right of way and from someone you want to take the right of way, etc. There's a lot of intelligence in this process. There are universal driving rules, and there are non-universal driving codes that vary from country to country. Mobileye has become a company that is a focus of all the technologies that are critical for autonomous driving."

Mobileye has 600 employees and a $9 billion market cap. How do you explain a company on such a scale having so few employees?

(Laughs) "$9 billion when the share is weak. At our peak, our market cap was $14 billion. Mobileye is growing all the time. I believe that we'll stabilize at 900-1,000 employees within two years in order to realize all our aspirations. The relatively low market cap is the result of our focus on software, which is what is good about it. It's true that there are other software companies with thousands of employees, and even tens of thousands, but the work we're doing can be concentrated in a relatively small number of employees. What enables us to make progress is the data we are accumulating, not the number of employees. Mobileye's products have been installed in vehicles since 2007, and the preparations for commercializing those products began in 2004. That means that since 2004, 27 auto manufacturers have been transmitting data to us for developing algorithms, something that is absolutely unprecedented. That's what distinguishes a product that is a nice demo from a product that you can rely on. For example, false braking can occur in a demo product. False braking is a catastrophe even if it happens once a week. The driver will immediately return the vehicle to the manufacturer. You have to achieve near perfection."

Wait a minute - near perfection or perfection?

"There is no perfection There is near perfection, and it is very, very, very difficult to achieve. This isn't something you see in a demo, because in a demo, you don't see extreme problems."

Do you really believe that a completely autonomous car can be developed, including human intelligence, as you explained?

"The question is what revolution is needed: scientific or technological? If it's a scientific revolution, you can't say how long it will take, because a scientific revolution can happen one year from now or in another 50 years. I say that the science already exists; in other words, the scientific revolution has already taken place, but the technology doesn't exist yet. The technological revolution is what has to happen. We therefore believe that there will be an autonomous vehicle in 2021, not next year."

In that case, what is missing in order to complete the technological revolution?

"Where the sensory element is concerned, the technology is ready, so what's left is to take care of mapping and driving policy. In order for an autonomous vehicle to become an industry, rather than just a technology, you have to build a business, and that means mobility on demand (MOD), like Uber. Uber's autonomous vehicle will be much cheaper, because the driver's work is half of the entire cost. In order to create a business, the price of such a service is very, very important, and the mapping element becomes the Achilles heel. This is a difficult logistics problem. If it isn't solved, the price of the service will be very high, and perhaps not worthwhile. Secondly, where driving policy is concerned, the trial vehicles currently existing are very simple in the way they manage complex situations. As soon as the road situation becomes complex, the vehicle is unable to cope with it, and the driver has to take over."

Let's get back to Israel's dominance in this industry. What is your opinion about the cooperation between Gett and Volkswagen?

"This complements what Mobileye is doing; it doesn't compete with it. Auto manufacturers like Volkswagen have realized that the auto industry is changing. In addition to the focus shifting to software, the definition of mobility is changing, and the auto manufacturers are therefore adapting themselves to these changes. Among other things, they are investing in cooperative ventures like that of Gett and Volkswagen. Gett represents the same thing that Uber represents - MOD - so it's natural for auto manufacturers to invest in it, just as General Motors invested in US company Lyft, which operates a travel sharing platform.

"In the future, the auto manufacturers will actually be companies providing mobility services. In contrast to the current situation, they will not be disassociated from the consumer as soon as they sell him a car. Just as Apple and Google continue to be part of our lives whenever there is a software update in their operating systems, Android and iOS, the auto manufacturers will maintain their connection with the consumer."

Let's move several years ahead. After we get through all the obstacles, regulators, and technologies standing in the way of a completely autonomous vehicle, will owning or using one according to need be economically worthwhile for the end consumer?

"The answer is complicated. Assuming that there will be an autonomous vehicle, two economic sectors will arise. The first will be MOD in urban areas, in other words, Uber without a driver. It will be very cheap for Uber itself to operate such vehicles, because it will save the cost of the driver, and the price of the journey to the end consumer is therefore likely to be worthwhile for the end consumer. The second will be vehicle ownership through either buying or leasing. In the case of leasing, the consumer can buy a monthly subscription, and will be able to call a different car each morning to take him where he wants to go. In other words, instead of owning a constantly aging car, the consumer will constantly replace it. That's terrific. In the case of a purchase, since an autonomous car will be more expensive than an ordinary one, it can be purchased in partnership with someone, such as a neighbor, or even several neighbors. All the partners will use the car together, each according to his needs. In this way, the car will be constantly in use, instead of only 4% of the time, as at present. The car may be expensive, but its cost is divided among several consumers, and its use will become more efficient."

Let's finish on a personal note. 17 years have passed since Mobileye was founded. How does it feel to see your vision materializing and succeeding on such a scale?

"I have no time to rest, because there are so many challenges and threats. Mobileye is a small company in a God-forsaken place, Israel. We're a long distance from our market, and each of our successes brings new threats with it. We're therefore constantly thinking about how to innovate, expand, and devise the right strategy."

Still, was there not a moment when you patted yourself on the back and said, "We did it!"?

"I had one moment, only one, and that was the day of the IPO in July 2014, when we were on the trading floor before going up on the New York Stock Exchange (NYSE) balcony, and saw the value of the share. Then we rang the bell to start the day's trading. That was a supreme moment."

Wait a minute - more supreme than the announcement of cooperation with BMW?

"Yes. We were the biggest Israeli IPO of all time in the US capital market."

Let me understand this. It's a matter of money?

"It's not the money; it's the achievement. It's like holding the world record in the 100-meter dash, and you know that no one will break it in the next 20 years, if at all."

While we are on the subject of money, remember that Shashua is one of Israel's few billionaires. His personal wealth, which is the same as that of CEO Ziv Aviram, his partner in founding Mobileye, is $1 billion, including shares that have already been sold.

The author was a guest of Mobileye at the CES exhibition in Las Vegas, Nevada.

Published by Globes [online], Israel Business News - - on January 12, 2017

© Copyright of Globes Publisher Itonut (1983) Ltd. 2017

Prof. Amnon Shashua (center) at CES Photo: PR
Prof. Amnon Shashua (center) at CES Photo: PR
Twitter Facebook Linkedin RSS Newsletters גלובס Israel Business Conference 2018