Click Image to Enlarge
One of the features of the 2011 Audi A8 that shows how electronics advances at the Volkswagen Group are aligned with consumer electronics at large is the ability to view Google Earth on the 8-in. LCD multimedia interface that uses a NVIDIA graphics processor as part of the navigation offerings. Researchers at the Volkswagen Electronics Research Laboratory in Palo Alto and at Google and NVIDIA have been collaborating for the past few years on achieving such capability.
Yes, that guy in the backseat of the Audi A8 is reading the New York Times on an iPad. What is of particular interest is that the vehicle features a WLAN—wireless local area network—which is said to be the first application of the technology as a factory-installed option. Up to eight devices—from iPads to PCs—can be connected to the Internet. The system uses UMTS (Universal Mobile Telecommunications System) and there is WPA2 encryption to assure data security.
"Of course we have the opportunity there to think about future technologies; that is one of our core competencies here. But we need to develop from the idea to the product--that's the intention and the goal." Dr. Burkhard Huhnke, executive director, Volkswagen of America's Electronics Research Laboratory, is talking about what he and his colleagues are doing at the facility based quite a long, long way from Braunschweig, Germany, where Huhnke received his doctorate in engineering in 1999. He suggests, however, that the way of advancing technology in Silicon Valley is why Volkswagen decided to site an electronics development lab in Palo Alto.
Although one might argue that the pace of development in the auto industry is significantly behind that of the consumer electronics industry--think of a cell phone of 10 years ago, a large clunky thing that could make phone calls, more or less, and nothing else compared with a smart phone of today; think of a car of 10 years ago and a car of today. Seems like the phone has far outstripped the car, doesn't it?
But Huhnke disputes that: "Yes, the car has four wheels and a steering wheel like it did before, but the architecture, the electronics architecture, is entirely different. Just think about the camera system used in cars today. Ten years ago the power of the algorithms, the speed of the network, and the systems integration capability wasn't at all capable." He also points out that the level of safety technology, predicated on electronics capability, is far more advanced than anything from a decade ago. He cites, as an example, the lane departure warning, blind spot detection, and emergency stopping capabilities on the Audi A8. "This was not available because it was not possible to integrate it into the car because the sensors weren't available or cheap enough."
The researchers at the ERL work in specific time frames that go out to 20 years in the future. Closest in, 0 to 3 years, is the work that is being done for series production vehicles. The predevelopment work is done 3 to 5 years out. Huhnke explains that a decision to implement a particular technology in a vehicle must be performed by the company board at about the 3-year point; ERL personnel work on the demonstrators and concept cars that are used in the presentations to the board of the technology. If the decision is made that the technology will be deployed, then the development work is done either by ERL personnel or it goes to a vendor. Then there is the work that is done 10 to 20 years in the future; often this work is done with university partners from schools including Stanford, Berkeley, and MIT. Huhnke says that working with university personnel is mutually beneficial, as the ERL personnel have the specific real-world experience and the students and professors have great theoretical knowledge. He suggests that it is important for car companies to work with universities, that there should be more than just monetary sponsorship involved. In fact, he says that the ERL engineers have a mentorship program with the university partners.
Volkswagen's automotive electronics initiative began in a big way in 2000. And the rationale for this need for understanding is simple: "If you want to assure the highest quality and reliability, you need the knowledge in-house." Given how rapidly the proliferation of black boxes occurred&mash;he points out that the first-generation VW Phaeton has more than 50 processing units, and it came out in 2002--this was essential. (The ERL was actually founded in August 1998 with three employees in Sunnyvale, CA.)
"Car manufacturers have long been steered by mechanical experts, but that's a huge risk because many of the innovations come through software developments and electronics," he says. And while they often work in parallel with suppliers, Huhnke maintains that it is vital for OEMs to maintain electronics competency, to not outsource the know-how: "If you just let your suppliers do it, you are steered by them." Engineers and researchers are keeping their hands on the wheel.
One of the big differences between the consumer electronics industry and the auto industry is the cost issue. This is the case for several reasons, although the primary one is that when you take into account the volumes of smart phone handsets or iPods, they are far in excess of typical automotive volumes. In addition to which, there is a big difference in the cost sensitivity of those who must have the latest consumer electronics product and those who are also paying for everything from engines to seat coverings to even paint. So on the one hand, auto is in the position of following the consumer electronics industry (i.e., they provide the economies of scale, which drives the cost down), but they must do what they can to provide similar functionality and capability.
One upshot of this is work with suppliers whose gear is more common to gaming systems than automotive systems. For example, in January it was announced at the Consumer Electronics Show that graphics processing units (GPUs) from NVIDIA would be integrated into Audis equipped with the vehicle manufacturer's third-generation multimedia interface (3G MMI). What's more, the interface system has been engineered so that it permits upgrades, thereby increasing functionality in a way characteristic of electronics but not typical of mechani-cal devices. Huhnke says they started talking with NVIDIA about working together on this program five years ago.
While Huhnke and his ~100 colleagues at the ERL embrace the Silicon Valley culture, there is one place where the two have a sharp difference. Huhnke says that while venture capitalists accept a one-in-20 success rate, that's something he can't afford. He cites a research study* that shows in new product development, of 3,000 raw ideas, 100 become exploratory projects, 10 of those well-developed projects, two full-fledged products, and a single success. Which makes him well aware that their development methodology must be one with a higher success rate. Much higher. "That would be one idea a day for 10 years," he says of the norm. "We could not afford that expense."
*"3,000 Raw Ideas = 1 Commercial Success" by G.A. Stevens and J. Burley; Research Technology Management; May/June 1997
Here are some of the key electronics initiatives being undertaken at the Volkswagen Electronics Research Lab in Palo Alto: