By Mark Shortt
Jim McBride, senior technical leader for autonomous vehicles at Ford Motor Company, points to a pair of vehicles prominent in 1960s science fiction—George Jetson’s car and the Batmobile—as evidence that the idea of self-driving cars has been around for a long time. “We just never had the wherewithal to actually do [anything about] it,” he told D2P in a phone interview.
That began to change in 2001, when the U.S. government decided it wanted to automate vehicles as a way of protecting the lives of military personnel on the battlefield. Three years later, the Defense Advanced Research Projects Agency (DARPA) launched the first DARPA Grand Challenge, a prize competition for autonomous vehicles, along a 150-mile route in the Mojave Desert. The series of challenges, the most recent of which took place in 2013 as the FANG (Fast Adaptable Next Generation Ground) Challenge, was open to the public to speed up the process of developing and building self-driving vehicles.
“We (Ford) participated in what was called the DARPA Grand Challenges because we thought if we looked at some of the new sensing and computing and algorithms that were becoming feasible, that we could probably translate those into making production automobiles not only more convenient, but safer,” said McBride. “Safety was absolutely the number one motivating factor. And once we did the Challenges and figured out that making a car self-driving was possible given the new wave of technology, all these secondary applications became obvious, like mobility for elderly or disabled people, or reducing urban congestion, or improving fuel economy. But it was always motivated by ‘What can we do to make cars safer?’”
Bobby Hambrick, CEO of AutonomouStuff, a supplier of systems, equipment, components, and services for the autonomous vehicles market, echoed McBride’s sentiments, saying that the biggest single factor driving the development of autonomous vehicle technology is the need to reduce the number of traffic accidents and fatalities.
“Thirty thousand people a year are dying, just in the United States alone,” Hambrick said in a phone interview. “That’s the equivalent of a 747 crashing every single day, and so that’s a major problem. If we said today that we were going to introduce a technology that was going to kill 30,000 people a year, that would never, ever be allowed in a million years. So it is a major thing.”
Technologies intended to protect occupants are now getting a boost from the market as the National Highway Traffic Safety Administration (NHTSA) revises its New Car Assessment Program (NCAP). Historically, the NCAP system rates a car as a 2- , 3- , or 4-star vehicle based on its ability to protect occupants after an accident. Now, as NCAP labeling begins to shift toward the prevention of accidents, a vehicle will need to offer automatic emergency braking in order to receive a top star rating. It will also need to have lane departure warning and blind side detection. “That’s going to drive it [autonomous driving] from the customer’s side, from the demand side of it,” said David Andrea, executive vice president of research at the Center for Automotive Research (CAR).
“If you look at the evolution of cars over the last decade, there’s been more and more driver convenience and safety features added to the vehicles—everything from adaptive cruise control to roll stability, to lane keeping assist, emergency braking, and things like that have continuously been rolling out on the cars,” said Ford’s Jim McBride. “So there has been an evolution in the technology from the perspective of the customer. But when you get to full autonomy, that’s sort of a paradigm shift where you’re not asking the driver to do anything whatsoever, and the car is assuming the whole responsibility. And that’s more revolutionary than evolutionary.”
Josh Hartung, CEO of PolySync, a software company building an operating system for developers of autonomous vehicles, believes today’s fervor around autonomous cars comes from a variety of perspectives. Autonomous vehicles create not only life-saving benefits, but also the environmental benefits of running fewer vehicles with higher utilization, he said. As a result, fewer vehicles are parked in cities, and people end up spending less time in traffic.
“On highways, it leads to much more efficient utilization of resources and transport of goods,” Hartung said in a phone interview. “Autonomous trucks can run continuously throughout the country, making it cheaper and faster to get things from point A to point B, and that has a ripple effect across the entire industry, every supply chain. Autonomous is one of those technologies that is exciting on so many fronts that it’s gained a life of its own. It seems like something that introduces an entirely new world, and I think that’s why it’s captivated the imaginations of so many people.” For more on PolySync’s operating system, see Building a Future of Software-Defined Vehicles.
Given all the potential use cases and potential benefits to society, McBride sees a strong push to make sure that autonomous vehicles mature and, ultimately, hit the road. It’s coming from not only automakers, such as Ford, but also from customers and a number of governments and regulatory agencies that also recognize the potential. “So I think there’s a very favorable outlook moving forward,” he said.
Contract Manufacturers See Opportunities
Although most of the activity around self-driving cars is still in the product planning stages, some contract manufacturers, suppliers, and job shops deep in the supply chain are not only well aware of the trend, but see it as an opportunity to apply their problem-solving capabilities to a new application. Two Wisconsin-based contract manufacturing companies, Strattec Security Corporation and DuraTech Industries, say they have capabilities relevant to manufacturing autonomous vehicle components and are already making parts for autonomous driving applications.
Strattec Security Corporation is a Milwaukee-based manufacturer of mechanical, electro-mechanical, and electronic security and access control products for automakers. The company designs, develops, and manufactures mechanical locks, electronically enhanced locks and keys, and ignition lock housings, as well as automotive access control products that include latches, power sliding side door systems, and power lift gate systems. DuraTech, on the other hand, offers printed and in-mold electronics, in-mold decorating, and membrane switch assembly, among other processes, at its headquarters in La Crosse, Wisconsin.
Strattec Sales Manager Warner L. Jackson said in an email to D2P that his company has been awarded business from a major automaker for a product that will be used within an autonomous driving system. “We can’t divulge any details yet because the vehicle has not launched, but it is a safety-related device,” he said.
Meanwhile, Paul Hatlem, sales manager at DuraTech Industries, said via email that DuraTech is currently making heating circuits that are used to clear snow and ice from panels located in front of radar for adaptive cruise control and autonomous braking. He added that DuraTech sees autonomous driving applications as an opportunity in the future because DuraTech’s processes “will allow designers more flexibility in 3D design and sensing while reducing weight and space requirements.”
“Like conventional copper wires, our printed or in-molded electronic circuits (wires and components) could tie into any vehicle—autonomous or traditional,” he added. “We know there are companies that could use printed or in-mold electronics to save weight and space over conventional electro-mechanical components. Anyone looking to make cars lighter or smaller could benefit from our technology.” For more on Duratech’s technology, see Wisconsin Firm Advances Manufacturing of Flexible Circuits.
Strattec’s Jackson anticipates that as vehicle product planning moves from the concept stage to the implementation stage, there will be a natural progression of activity down the supply chain. Strattec has identified multiple areas of the vehicle where its product and manufacturing capabilities can be utilized.
“We believe that autonomous driving not only changes the way the vehicle behaves, but also the way people will interact with the vehicle,” he said. “The vehicle operator will not be as focused on the driving activity and will have greater ability to interact with the passengers. The structure and layout of the cabin will see significant changes, which may provide opportunities for Strattec to introduce new products.”
Jackson also noted that autonomous vehicles are about more than just the software that controls the steering, braking, and acceleration. As the overall structure and content of vehicles changes, they will demand new mechanical and electro-mechanical systems that are instrumental in providing convenience and entertainment to the vehicle occupants, he said. He believes the development of self-driving vehicles will bring about some changes within the supply chain.
“We believe there will be more collaboration between OEMs and between suppliers,” Jackson said. “The complexity and cost of the new systems will demand that companies seek partnerships that can expedite the design and manufacturing phases and can bring greater economies of scale.” For more on Strattec, see Strattec Grew Up with and Evolved with the Car.
At the upper end of the automotive value chain, there’s competition not only among automakers, but also between automakers and new tech entrants in what is often described as a new mobility ecosystem.
“We’ve heard a lot of the headlines around Uber, and Google, and Apple,” said Daron Gifford, an analyst at consulting services firm Plante Moran in Detroit. “My suspicion is that they’re all trying to figure out whether they just want to build their own vehicle, or they just want to apply the technology to somebody else’s vehicle, and then be the technology holder, in that case. The OEMs are competing directly with them because they’re trying to develop not only their vehicles of the future, but the same kind of technology capability around the autonomous vehicles.
“The challenge with the existing OEMs, even though they’re large companies, is that their cash reserves and capital pale in comparison to Google and Apple,” he continued. “It’s just how the market has worked in the past. Apple’s got probably a $600 billion or $700 billion dollar market value, $200 billion in cash reserves, and nobody in the auto industry is even close to that.”
Automotive Industry Begins 2017 in Good Health
Dave Andrea, executive vice president of research at the Center for Automotive Research (CAR), said that the U.S. automotive industry is in good health from the standpoint of total market demand on sales and production. He used the word “hypercompetitive” to describe the industry from the vehicle manufacturer level, down through the supply chain—including materials suppliers—and the dealership network. “Everybody really got their balance sheets cleaned up, and reorganized through the financial crisis in 2008 and 2009, so these are really the top players here,” he told D2P in an interview.
Andrea described the automotive market as “dynamic” due to strategic challenges posed by emerging technology, as well as uncertainty about where the economy is headed and how it will be affected by the new administration’s positions on trade and regulations. He also characterized it as being defensive, meaning that the industry “still doesn’t have the benefit of the doubt from Wall Street” that automakers have restructured enough to be able to survive through the next downturn with a positive return on capital.
“In the U.S. market, we’re plateauing out here. You see, from the publicly traded firms, a lot of pressure on their price to earnings, multipliers that Wall Street is applying. Many of them were cut to either a ‘market perform’ or ‘underperform’ after the election, just given the uncertainty around the regulatory, trade, and tax environment for the next four years,” he said.
A 2015 study that CAR conducted for the Alliance of Automobile Manufacturers bears out the powerful impact of automotive manufacturing on the U.S. economy, particularly as a creator of jobs, associated compensation, and tax revenue. The study showed that auto manufacturers, suppliers, and dealers support more than 7 million private sector jobs in the U.S., and that the automotive industry is responsible for $500 billion paid in annual compensation to employees. Of the more than 7 million jobs supported by the auto industry, 2.44 million are created by automakers, 3.16 million by suppliers, and 1.60 million by dealers, according to CAR.
Andrea said that typically, one assembly plant job creates an additional six to seven jobs in the rest of the economy, including throughout the supply chain and at grocery stores, restaurants, and other places where employees spend their paycheck.
How Autonomous Vehicle Development is Impacting the Supply Chain
The promise of innovation is leading many automakers to make investments in new automotive technologies and engage in pilot programs to test them, said Plante Moran’s Daron Gifford.
“They’re really looking for suppliers to help them with innovation, he said. “What can they get from them to apply toward their model? A lot of the OEMs, as large as they are, realize it’s not really possible for them to be able to take all these pieces on their own and build it on their own. So there’s going to be a lot more partnering, probably some joint ventures, where you can get more of that early development capability, versus just bidding out awards of parts to suppliers.”
The trend toward development of more advanced driver assistance systems (ADAS) and autonomous vehicles is impacting the automotive supply chain in a variety of ways, including consolidation. Many companies are seeking to broaden their capabilities through mergers and acquisitions of suppliers with expertise in the software and electronics needed for development of self-driving cars. A recent example was Samsung’s announcement in November that it would acquire Harman for $8 billion—a move that combines Samsung’s expertise in connectivity technologies and semiconductors with Harman’s capabilities in connected car technologies. Another high-profile acquisition was ZF’s acquisition of TRW in 2015 for more than $12 billion.
A 2016 report by PwC, “Consolidation in the Global Automotive Supply Industry,” states that “the race toward fully autonomous vehicles will fuel more megadeals for ADAS suppliers. Already valued at $31 billion, the market for ADAS technology is likely to triple in size by 2025, as OEMs move beyond function-specific capabilities, such as adaptive cruise control, to engage in full vehicle automation.”
Bringing together all the parts and components required for ADAS and autonomous driving systems is a challenge in itself.
“They’re looking at calibration and how important that piece is because we can have the bill of materials with the 3,000 to 5,000 individual parts, but how do those parts all come together?” Andrea said. “What are the software and the algorithms, and the calibration of all these components together that really deliver customer satisfaction and really deliver the differentiation between what makes a General Motors vehicle or a Toyota vehicle?
“It really depends on the calibration and the software development that makes those shift points a Ford; that makes the rev curve a General Motors vehicle,” he continued. “But the industry needs to do those kinds of changes to be able to get the scale to be able to keep vehicles affordable as they deliver the upper end of the fuel efficiency range, with very, very small levels of emissions, in the safest vehicles that we’ve ever put onto the streets.”
As it becomes clear what new skills are needed to develop autonomous vehicles, both OEMs and suppliers will need to adapt by bringing more software engineers, those skilled in integrating hardware and software, into the fold. “When you ask the suppliers what are the largest number of open engineering slots that they have, almost always, they say that they’re for software engineering,” said Andrea. “Software engineers are in the greatest demand through the entire supply chain.”
One of the challenges facing OEMs, such as Ford, and their suppliers is that the software is highly dependent on the hardware that it’s running on. “It’s not simply just plug-and-play some software,” said Ford’s Jim McBride. “The software is dependent upon the assumptions about what information is coming into the perception system, for example, or what information you might have by way of prior maps or knowledge of the area, or communication to and from the car, and things like that. So it really is a tightly coupled system, and you can’t just put software or tech on the car independent of the hardware.” See PolySync’s approach to solving this challenge in Building a Future of Software-Defined Vehicles.
McBride acknowledged the need for engineers who are well-versed in systems integration. The question is whether the OEMs are going to own the whole system integration themselves, or whether they’re going to need full-service suppliers, he said.
“In the past, a car could have parts from different people—say, radar from Delphi, and brakes from Conti [Continental], and each of these systems—maybe the airbags, or the back-up camera, or the blind-spot warning, or the cruise control—operated independently,” said McBride. “It was easier to mix and match the parts. But on an autonomous vehicle, all those things only work if they communicate with one another and share data, so they have to be tightly integrated. And that makes it more of a challenge to sort of randomly pluck suppliers and put their pieces together on a car. I’m not saying that you couldn’t have different suppliers, but you’d have to have a lot more systems integration, and so that lends to either the OEM doing it by themselves, or some of the traditional suppliers trying to be a full-service provider of a suite of technology.”
What does the automotive supply chain need to do in order to be ready to serve the needs of the OEMs who are developing autonomous vehicle technology today? Some of the challenges ahead highlight the importance of being able to work well together, according to McBride.
“The supply chain needs to understand what the requirements and specifications are—how far does a sensor need to see, and at what resolution?” he said. “And what frame rate do they have to update the data? And given that we’re still out there trying to drive and experience all the scenarios that are possible in the world, we’re still gathering data on what it takes to drive every possible scenario that humans normally see. In some instances, some of the specifications are a bit of a moving target. So for the suppliers, they’re going to have to try to work with the OEMs and technical experts to make sure that they cover the bases with regard to having the fidelity of data that’s going to be needed to pull this feat off.”
Autonomous vehicle development also impacts what automakers look for in a supplier, McBride said.
“We believe that any single sensing mode that you would typically find on a car today has some weaknesses in addition to the strengths it has,” he said. “So our premise is that we’re going to exploit multiple sensing modes in parallel, and multiple algorithms in parallel, to achieve the level of robustness that we would need. And so the suppliers would also have to operate in that vein. If they have one sensor, they have to have complementary sensors that fill in and strengthen the picture.
“Suppose you had a supplier that only did cameras and radar, and they had no LiDAR, or they had no inertial navigation system, or they had no interface with a mapping company,” he continued. “They’d be hard pressed to provide a full service solution unless they partnered with people that could put those other pieces in there and make sure that they covered all the bases. Or the OEM would have to assume ownership of the whole process.”
Another challenge has to do with hardware. Many of the autonomous test vehicles currently on the roadway rely on hardware that isn’t production-grade because there hasn’t yet been a market for some of the sensors and high-powered computers that are on those vehicles, said McBride.
“The supply chain has kind of been caught a little bit behind the ability to solve this problem,” said McBride. “So from a hardware perspective, it’s going to be a challenge to, say, develop supercomputers and LiDARs, and all the other new things that go on the car, en masse, at a price/volume curve that is economical for everybody. I don’t think it’s an insurmountable challenge, but it’s just that we’ve all come to the conclusion that we can make cars that drive themselves, and that conclusion kind of came before the supply chain was ready to give us the parts to do so.”
The 5-to-10-Year Outlook
When asked what he thinks the automotive supply chain will look like in the next five to 10 years, McBride said he doesn’t believe that cars, as they exist today, will necessarily be replaced. Instead, he believes that consumers will have a host of new mobility options, such as ride hailing and sharing services. In that case, theoretically, cars would operate more hours per day.
“You’d probably end up replacing those cars more frequently than a personal ownership vehicle, which tends to sit in a person’s garage for 10 or 11 years before they buy a new one,” he told D2P. “But from the company’s point of view, you’d have additional revenue streams that had never been available to us. Right now, transportation as a service is, I think, roughly a $5 trillion a year enterprise, and the car companies have virtually no share in that. So, by moving into the mobility space, you can start to generate revenue from providing mobility as a service.”
Plante Moran’s Daron Gifford said that he tries to encourage clients to really be aware of what’s occurring and then try to identify early on what role or position they’re going to take in this new automotive—or mobility—ecosystem.
“From a supplier’s standpoint, we tell our clients, ‘You can just watch this, because you’re not sure where it’s going, and become a follower, or you can try to get involved in it and take a leadership role,” he said. “There’s some risk with that, but the rewards are pretty great, and the risks around just following and reacting could be pretty great in a negative way. We ask them who they want to get themselves engaged with, and what core competency are they going to bring to this vehicle of the future?”
Mark Langlois contributed reporting to this article.