Volkswagen is taking great strides in making the roads safer and remove the dangerous fun from driving by developing fully autonomous vehicles. I had a chance to talk to Dr. Burkhard Huhnke, director of the Volkswagen Electronics Research Laboratory (ERL) about the future of the Volkswagen Automotive Innovation Laboratory (VAIL) at Stanford University and how the technology developed there is being integrated into Volkswagen Group vehicles. You may be able to buy a real-life K.I.T.T before you know it. It probably won’t be a Pontiac though.
You may recall that Volkswagen was the first team to complete the DARPA Grand Challenge in 2005 by having a fully autonomous Volkswagen Touareg SUV (his name was Stanley, btw) drive 132 miles through the Mojave Desert. Then for the 2007 DARPA Urban Challenge, a VW Passat Wagon took second place behind Tartan Racing team from Carnegie Mellon University in a 60 mile urban course. But those two challenges are nothing compared to what’s on tap for next year: Pikes Peak in an autonomous Audi TT-S.
CrunchGear: How about we start with an overview of the project.
The VAIL is an initiative that we started here from our electronics research lab in Palo Alto to initialize automotive center at Stanford. We got the smartest people from all over the world in one place. We already had experience with the DARPA Grand Challenge race success. We won the desert race with the autonomous driving cars together with the Stanford racing team. Then we repeated the success in the Urban Challenge in 2007.
So we thought how can we bring each competence together, create a platform for a convenient automotive industry and Silicon Valley partners to look into the future and find some solutions for the mobility, safety and environmentally-friendly problems, which we have to face together.
The Volkswagen Automotive Innovation Laboratory funded a building, which will be dedicated and opened tomorrow (10/24/2009). We’ll fund the research projects with $750,000 per year for the next five years.
CrunchGear: How is Stanford involved in the project?
We expect first that we collaborate with the Stanford genius together on the solutions that we create for the future. We divided the VAIL topics into three major themes
One is how do we want to drive autonomously in the urban environment. We have so many problems that we have to solve like direction recognition, lane warning, lane detection, but all these obstacles have to be observed by camera systems in the urban environment, which is one of the biggest challenges.
What we would like to create for all of these three topics, and I’ll come back to the 2nd and 3rd, is initiatives. The vision for urban driving is to drive from San Francisco to Los Angeles autonomously and that involves traffic jams and obstacles. What can we do to provide to our drivers with? They want to text while driving and we have to make that safe.
We already established systems in our current models like blind spot detections, adaptive cruse control but what’s the future for that? Can we already allow them to press the auto button to start texting? So that’s the vision. Drive along the highway and allow the driver to do something other than driving.
The second vision is up to the limits together with the driving dynamics department at Stanford. If you come to a critical situation, you generally have to take the control over from the driver.
Let me explain it to you this way: If you brake your car without control, you will definitely hit the obstetrical in front of you. Can you imagine that we can try to find the emergency parts for the driver and turn the car into a different path to avoid any impact any with the obstacle? It would be a smart system that acts within the limit’s range of the car that would find an emergency solution. The driver would never be able to do this because there is too much information within the milliseconds.
The idea for up to the limits is to look at the best drivers to give us super examples how to react in critical situations. So we looked at a German rally driver that drives up Pike’s Peak. It’s a really critic race, but it’s a challenge for us to try to understand what makes this rally driver so special and get this knowledge into the car.
We built a next generation autonomous car that’s able to drive like a rally drive and drift into the curves. It drives even in extreme situations. We think we can actually drive up hill to Pike’s Peak next year. That’s the second project.
The third one is with multitasking. We actually sponsored a project seen in the NYT on what people are really able to multitask or are they not really able to focus on one task and fulfill that. It’s really an interesting study. But for us it’s really interesting because we would like understand what the perfect interface between humans and machines. Do you believe in your car? Do you believe that if you press the autopilot button that the car would actually do what you would expect. Or do you have blue screens like we experience everyday – I don’t want to blame Windows – but with our operating systems. We have to prepare the highest reliability for systems. But what do you expect in these relationships between the car and the driver in the future.
These three founding purposes are the main intentions at VAIL.
http://blip.tv/play/grExganxGwA
CrunchGear: What type of testing has the Pike’s Peak vehicle seen so far?
We actually tested that at Salt Flats to have the open area available to be able drive up to the limits without any risks. It sounds crazy but we needed the smartest people actually at Stanford to prepare to the algorithms because you have to control systems again in the in-stable mode and the few years ago we thought it was impossible. But we did test already and we created a movie to show that’s possible and we think we can go one step farther and do this crazy race.
Again I told you that it’s rally style driving. It’s really incredible; up to the limits.
CrunchGear: One of the reason they use Pikes’s Peak is because of the environmental conditions. How does the altitude affect the vehicle?
We have the best system actually to use for that in the Audi TT-S and it’s working perfect. We tried it already at Pike’s Peak in slow mode to figure out what type of problems we might have in that environment.
CrunchGear: What else has changed in the Audi TT-S
If you saw our Junior VW Passat Wagon urban environment, we had many roadblocks to avoid and accidents with any person in that car. We don’t need the rack on the roof anymore. We don’t need big sizes for that at all. It looks almost like a stock car. You would only see three antennas on top of the roof. That’s all. We fit the computer in the trunk. We really want to show off that we’re coming closer to an actual product. It’s not too far away.