Intel and Mobileye: Building an Ecosystem for Automated Driving

By Charles King, Pund-IT, Inc.  December 13, 2017

The tech industry is no stranger when it comes to future-focused products and trends. That they tend to generate more heat than light can be blamed on systemic hope and the desire to find and capitalize on the Next Big Thing. Or maybe call it an optimistic reflection of the industry’s remarkable history of delivering transformational products and services.

Of the future-focused technologies currently in the spotlight, it would be hard to find one with more transformative potential than automated driving, including autonomous or self-driving passenger cars and commercial vehicles. Why so? Partly because of the sea change in behavior they could spark among individual consumers and businesses, especially those located in urban areas plagued by traffic jams, long commutes and other woes.

But automated driving is also worthy of examination due to the variety, size and seriousness of projects being pursued by companies inside and outside of Silicon Valley. Those range from IT vendor stalwarts and early stage start-ups to Big Five tech giants and high-flying ride-sharing companies to mainstay players in the automotive and other industries.

In other words, the sheer volume of effort and investment in automated driving, along with advances in related areas, like digital mapping, mobile and wireless, sophisticated edge computing, endpoint sensors and high-performance networking makes it a subject worth investigating. Intel recently hosted an Analyst Day event that highlighted the company’s efforts and advancements in these areas that is well-worth close consideration.

Building on Mobileye’s foundation

Intel’s efforts center on its Automated Driving Group (ADG) which accelerated even more after the company acquired Mobileye, an Israel-based autonomous driving pioneer, in a $15.3B deal that closed on August 8th. Doug Davis, the Intel SVP who runs ADG, which now reports into Mobileye’s co-founder and CEO Prof. Amnon Shashua, opened the Analyst Day events by detailing high-level automated driving concepts that Intel targets.

Davis underscored the practicality that is so central to the company’s approach—the need for emphasizing commonsense designs and identifying, then solving tough engineering problems. Those are certainly all within Intel’s purview but Davis also spotlighted issues that are relatively unique to the automated driving market. For example, the high level of technological sophistication required to develop solutions, the safety-critical nature of these products, and the value of relationships with established auto industry players makes it difficult for newcomers to make much headway.

That runs contrary to Silicon Valley’s customary wild love of innovative start-ups, but the issues aren’t problematic for Intel’s Mobileye/ADG. Mobileye was originally co-founded in 1999 by Prof. Shashua, the company’s CEO and its CTO (and an Intel SVP) who has profound knowledge of and experience in computer vision technologies.

Today’s success and future opportunities

Mobileye began shipping commercial vision technology products in 1999, and today its portfolio encompasses Advanced Driver Assistance Systems (ADAS) solutions, such as passive alerts, like Lane Departure Warning (LDW) alerts and Forward Collision Warning (FCW), and active safety systems, including Automatic Emergency Braking (AEB) and Adaptive Cruise Control (ACC).

How has it fared competitively? Quite well. Today, Mobileye works with virtually all of the major automaker OEMs. In 2010, the company sold about 182k of its machine vision chip units and its solutions were used in 36 car models built by seven OEMs. By 2016, Mobileye chip sales had ballooned to 6.0M units, and in 2017 its offerings were being implemented in 313 car models from 27 OEMs.

Mobileye’s business continues to grow at a healthy rate. The company enjoyed fourteen design wins in the first ten months of 2017 compared to 12 wins in all of 2016. Intel’s Davis noted the company’s sustained win rate (with 22M ADAS systems shipped to date) and high profitability in his remarks.

Those points are obviously important to Intel, especially considering the $15B cost of the acquisition. But more importantly, Mobileye’s deep expertise in such potentially dynamic markets and its broad acceptance within the automotive industry offer a solid foundation for supporting current and future initiatives and efforts.

Broadening the ecosystem

Other Analyst Day sessions by Intel and Mobileye executives, including Prof. Shashua, dug into related subjects. Those included a deep dive on core Mobileye technologies, discussions of its semiconductor development efforts and strategy, the key role data center infrastructures play in automated driving and an overview of the Intel ecosystem assets that are crucial to the success of Mobileye solutions.

Why crucial? Because much like other emerging technologies, such as the Internet of Things (IoT), success in automated driving will require significant development of and investments in core IT infrastructures and services. Despite those issues, the media and other interested bystanders tend to focus attention on high-profile early movers in autonomous vehicles, like Waymo/Google, Tesla and Baidu, and interested automakers, including Ford, GM, Renault-Nissan and Daimler.

That enthusiasm is understandable and initial self-driving vehicles are expected to be commercially available in the early 2020s but these vehicles will not see significant penetration on our roadways for some years. In other words, focusing mainly on today’s self-driving vehicles is a bit like obsessing over amazing new smart phones that, son of a gun, won’t be supported by mobile networks for 5+ years.

Intel clearly understands those points, including the dichotomy of commercial interests, self-promoting players and fickle public attention. In response, the company is following a sustainable path and bringing along partners and customers for the ride. The key to Intel’s success lies in developing a broad ecosystem of technologies, solutions and services that are applicable to most every phase of automated driving and meet the discrete needs of businesses focused on those markets.

Mobileye is clearly the tip of the company’s formidable strategic spear, but Intel-based silicon will play a huge role in the systems, storage, memory, networking and other infrastructure assets that keep autonomous vehicles dependably and securely on the road.

Safety and behavioral challenges

One of the most interesting Analyst Day sessions was led by Marjorie Dickman, Intel’s global director and associate general counsel for automated driving and IoT policy. Dickman dug deeply into an issue that several other presenters noted—the critical role of safety in self-driving acceptance and adoption. She emphasized that this focus on safety is top priority in Intel’s government advocacy on automated driving around the world. In fact, industry players that advance too quickly or at the expense of safety could severely risk public trust and the opportunities they hope to capture.

Complicating efforts to bring self-driving vehicles to fruition are patchworks of often conflicting federal, state and local regulations in the United States, as well as the European Union, the United Nations and numerous individual European and Asian countries. Some are also in flux. For example, in the U.S., the House of Representatives passed the bipartisan SELF DRIVE Act on September 6, 2017, while the subtly different Senate’s AV START Act passed a committee vote but is waiting for a full Senate vote, and reconciliation with the House bill.

It seems likely that other, more divisive political measures will overshadow adoption of a House-Senate reconciled self-driving bill in the last days of the 2017 Congressional session (which ends on December 21, 2017). Plus, you have to factor-in efforts by diverse interested parties, including cybersecurity, environmental and trucking industry representatives, and stakeholders lobbying on both sides for various preemptions and exemptions.

All in all, a bicameral bill covering automated driving in the U.S. is likely to be adopted by Congress sometime in early 2018. If passed, the legislation—together with the Department of Transportation’s recently updated guidance for automated vehicles—would create a much-needed uniform federal framework for self-driving vehicles.

What I found fascinating about Dickman’s presentation was how it clarified issues impacting and shaping the development of automated driving technologies and demand for autonomous vehicles that have remained relatively invisible or underdiscussed in more public venues. Intel deserves credit for its willingness to openly discuss these often-thorny points during the Analyst Day event.

Final analysis

From where I stand, autonomous vehicles and self-driving technologies could very well become the Next Big Thing so long as your outlook extends to around 2025. Before we get there, however, technology vendors and automakers have a lot of work to do, including addressing and solving central problems which are, to put it gently, often profoundly tough and complex.

Keeping those points in mind, I came away from Intel’s Analyst Day with respect and optimism for its accomplishments in automated driving. If the company can continue developing its ADG technology ecosystem, delivering the solutions its executives discussed and achieving the vision they detailed, Intel will be a key player in the transformational revolution that autonomous vehicles seem likely to deliver.

If that comes to pass, August 8, 2017—when the acquisition of Mobileye was completed—is likely to be remembered as the day Intel’s journey to automated driving and autonomous vehicle leadership stepped on the gas and moved into the fast lane.

© 2017 Pund-IT, Inc. All rights reserved.