To Bring Driving into the Infotainment Age, GM’s Palo Alto Office Melds Silicon Valley Fancy with Detroit Pragmatism
I visited GM’s Advanced Technology division in Silicon Valley on August 4, the same day the Detroit giant revealed its second-quarter financial results. The numbers were far better than one might have expected, given the automaker’s troubled recent history. GM said it brought in $2.5 billion in net income in the quarter, which was a bit less than it had earned in the previous quarter but nearly double the figure from a year earlier, back when the company was still mostly owned by Uncle Sam. The same day, GM said that its market share, after six years in free fall, has started to tick back upward.
Now, that’s encouraging data—but your mind is probably stuck a few sentences back, saying, “Wait, what? GM has an office in Silicon Valley?” Innovation, after all, isn’t a big part of the brand image at a company where the most iconic and profitable product is still the massive, gas-guzzling Chevy Suburban SUV.
But indeed, GM (NYSE: GM) does have such an office—inside a converted HP manufacturing facility in Palo Alto, on the same street with more conventional Silicon Valley players like WePay, mSpot, and Fry’s Electronics. GM doesn’t play up the facility’s existence in the press, and it took me more than a year to score an interview with its managing director, Byron Shaw. But once I’d toured the place and had a chance to quiz Shaw on his mission, I realized I’d picked a pretty appropriate time to visit. The situation at the office was—well, far better than one might have expected, given the automaker’s troubled recent history.
The biggest achievement for the 10-person outpost, which Shaw set up in 2006, is that it has managed to stay open for five years, despite the company’s bankruptcy, and broader turmoil in Detroit and the larger economy. “Surviving through the 2008-2009 downturn was key for us,” says Shaw. “We were able to maintain this operation without any cuts or loss of resources during that time. That was a big thing.”
But GM has a common-sense reason for keeping the Palo Alto office open, and it’s this: When you’re in your car, you’re not just driving anymore—you’re likely in “infotainment” mode, toggling between satellite radio, the GPS navigation system, cell-phone calls, and the like. Which means the components that make a car stand out in the marketplace are no longer made just from steel, rubber, or glass, but also from electronics and software. “The writing is on the wall,” says Shaw. “Look at what a car is today, and look at what it will be in 20 years. If you want to be competitive you have to have the best software and electronics in the industry, and you won’t get there if you don’t participate in Silicon Valley.”
Shaw’s squad doesn’t actually write software or build electronics. Rather, its job is to build relationships with all the people who do—so that when a new idea comes along that might make a future GM car better, the company knows about it.
Take Terminal Mode as an example. One staffer from Shaw’s office represents GM within an informal group of companies called the Car Connectivity Consortium (Honda, LG, Motorola Mobility, Nokia, Samsung, Sony, Toyota, and VW are also members). The consortium is working on a way to make the infotainment system in your car into an extension of your smartphone. The idea is to let you access the content and apps on your phone via the in-vehicle sound system and LCD displays, with all user-interface elements automatically optimized for use while driving. (Sorry, no Doodle Jump allowed.)
Terminal Mode is both a set of Internet-based standards for moving all this data around, and an emerging system for evaluating and certifying new apps to make sure that they don’t exceed driver-distraction guidelines. Shaw says that Silicon Valley is “pretty much ground zero” for the consortium’s negotiations over Terminal Mode. For GM customers, the benefits of the company’s participation will be pretty obvious: they’ll be able to tap the power and connectivity of their smartphones in their cars, but using simplified interfaces that automotive engineers have vetted as safe. “Maybe there are 15 navigation apps out there for your phone,” says Shaw. “What we will be able to do is certify that ‘These three nav apps have been tested and they meet GM guidelines, so you’ll get more support [in Terminal Mode] if you use them.'”
Shaw also has representatives on the board of GENIVI, a Bay Area-based group of companies developing an open source operating system for the in-vehicle infotainment (IVI) systems that Terminal Mode might connect to. Right now, pretty much every vehicle model and generation has its own unique IVI, which means developers must write the same basic navigation, music, Internet, multimedia, and communications apps over and over. The GENIVI companies want to create a write-once, run-everywhere environment similar to iOS or Android, so that developers can spend more of their time being creative. Of course, if GM ends up using GENIVI in its own vehicles, it will have the added challenge of differentiating its own infotainment systems in the marketplace. But Shaw’s argument is that “If you don’t participate, how can you not be behind?”
It all makes perfect sense to anyone from the software world. Which is why it probably would have been hard for a GM insider to start the Palo Alto office. Instead it took someone like Shaw, who has solid credentials in the auto industry but has spent most of his career in Silicon Valley.
Shaw’s personal history with GM is, however, complicated. He’d been a garage monkey since childhood, tearing apart every car or motorcycle he could get his hands on. And his very first job, after leaving MIT in 1990 with a BS in mechanical engineering, was in GM’s Harrison Division, where he was asked to help redesign vehicle air conditioning systems to use a new, environmentally safer refrigerant. But it wasn’t a happy experience. “I ran away screaming, and I swore to myself that I would never work at GM again,” Shaw says.
The problem, Shaw says now, was that he had been trained at MIT to try to understand a problem before solving it. But the ethic at GM was simply “build it, try it, test it until it breaks,” he says. “It wasn’t ‘model this, understand the physics, see if your refrigerant is chemically incompatible with your lubricant.’ There was a disconnect for me.”
Shaw says he felt much more comfortable in his next job, testing new engines in the central research labs at Daimler Benz in Stuttgart—and not just because he happened to have a second degree in German language and literature. “That was a great experience, because I was able to do exactly what I was doing at school, in an applied fashion,” he says. It was a huge contrast to the product development work at GM, where he says there was “very little stomach for experimentation.”
Eventually Shaw moved to California, where he got his PhD at UC Berkeley. His dissertation used computer models to explore ways of reducing hydrocarbon emissions from engines at cold-start. He did some work for a software firm—“The lure of being sucked into working for a software company in Silicon as a mechanical engineer is very high,” he says—but, unlike many of his peers at Berkeley, he decided he wanted to stay in the auto industry.
That was fortunate for another German firm, BMW, which, around 1998, was looking for someone to start a Silicon Valley office, as its competitor Mercedes had in 1996. “I was the first hire,” Shaw recounts. “A couple of guys came over from Germany, and we had an admin. We didn’t know what we were doing. I started doing prototyping. And that was just a winning formula for BMW. They really liked it.” The BMW office essentially turned into a production shop for high-tech show cars, Shaw says.
“What I learned from that experience,” he says, “was that if you’re going to take technology in its purest form, present it to another industry that embraces technology but doesn’t understand technology and is built around a consumer need, transportation, they don’t appreciate the beauty of a chip or a correctly written API. You need to put it in a context. And in the case of automotive execs, show cars is what they’re used to.”
But the prototypes led to some real products: Shaw’s facility built the first version of iDrive, the computer interface that’s now used in most BMW cars to control the audio, navigation, and climate systems. A skunkworks project led to the world’s first system for connecting an Apple iPod to a car audio system—again, standard equipment on most cars today. “We had a lot of fun there,” Shaw says.
Shaw left BMW in 2003. That’s roughly when GM began courting him to rejoin the company. “I had been talking back and forth with some people I knew at GM, and they asked me to come to Detroit. I thought I was going to, but when we got to the altar, I had to say, ‘Sorry, I can’t do this. I can’t move to Detroit. But if you happen to put up a facility in Silicon Valley, give me a shout.'” Shaw returned to California and took a detour into the solar industry, founding a company called Ready Solar that specialized in retrofitting residential rooftops for photovoltaic panels.
The shout from GM finally came in April 2006—and in some ways, it was a cry of pain. Conditions were bad inside GM. The company was losing market share to Toyota and other global competitors, and was hemorrhaging more than a billion dollars a month. Some media outlets were predicting the automaker’s demise. But it was ready to listen to Shaw’s argument that it needed a beachhead in Silicon Valley.
“GM wasn’t exactly printing money that year,” says Shaw. “It was a big deal, in the middle of that, to convince GM that they needed to do something bold in Silicon Valley. Out here we used to say all the time, ‘Look who’s not at the party—GM, Ford, and Chrysler.’ All of the multinationals—BMW, Daimler, VW, Audi, Toyota, Nissan—had a presence in Silicon Valley, all had embraced the fact that the spin cycle here is so fast that if you are not in the middle of it, you are not going to be able to lead in those spaces. Getting GM to do that was a big step forward.”
But this gig was nothing like Shaw’s BMW years. There are no show cars inside the big, open, startup-like Palo Alto office—just engineers figuring out how to apply Silicon Valley thinking to the driving experience. One of Shaw’s lieutenants, for example, is Frankie James, who got her computer-science PhD from Stanford studying HTML interfaces for the blind. Now she works on improving the human-machine interface inside vehicles. “You might as well think of drivers as being blind—their physical capacity for sight should be focused on the road,” Shaw explains. “But the touchscreen seems to be the baseline [user interface] going forward. So what technologies could you use to allow you to have touchscreens in the car that don’t require you to look at them?”
When Shaw’s people find vendors making software or gadgets that might work inside cars, they can put the products through their paces in the facility’s electronics lab. There’s even a garage where they can see how new technology might fit into actual GM vehicles. The process shows not just whether the technology is ready for the road, but also whether the vendor is ready to deal with a big, demanding customer like GM, says Shaw.
The next step isn’t to impress an executive with a smoke-and-mirrors demo, but to “get an internal champion inside the company who is going to start doing the real work, so that we can just be the instigators,” Shaw says. He says he learned at BMW that “You have to work soldier to soldier with the engineering staff up front, so that when the executive says ‘Hey, what do you guys think about this thing I saw in Palo Alto?’ they can say ‘Yeah, we’re working with them on that, everything is copacetic.’
In other words, there’s a lot of organizational politics and bridge-building to Shaw’s job. Investing, too—Shaw is a managing director at GM Ventures, and oversees the company’s corporate venture activities in Silicon Valley. But having a Palo Alto office is also about learning. “There are a lot of lessons learned that are not just straight tech-product evaluations, but sort of process and best-practice evaluations,” he says. “What is it that makes Silicon Valley such a hotbed of innovation, and how can we integrate that inside of GM?”
As product life cycles get faster, that may be a question of life or death for Detroit. It used to take 60 months or more to get a new car from the drawing board to the assembly line. Now the figure is more like 30 months, and companies like Toyota are trying to get it down to 12. The only way to move that fast, Shaw argues, is to incorporate a few common Silicon Valley practices—i.e., publishing your specs (as the Terminal Mode and GENIVI groups have done), using open-source code, and collaborating across corporate boundaries.
“It’s a question of how much of that can we embrace,” Shaw says. “We certainly feel that open source, collaborative environments, partnerships, a lot of these techniques can be injected into the industry.”
Of course, there’s a limit somewhere. Cars aren’t smartphones; consumers don’t want to be the beta-testers for machines that cost $30,000 and carry their families at 75 miles per hour. So part of GM’s job in Palo Alto is also to provide a reality check for Silicon Valley software developers, forcing them to think about what sorts of interfaces are truly safe and usable inside vehicles. Speech-synthesis interfaces that let you listen to your e-mail while you drive? Probably. Angry Birds? Probably not. Says Shaw, “That’s the light at the end of the tunnel for me—if we can find this melding where we are bringing our respective strengths to the table.”