I have been quite busy for the past few days as I looked for new
Volvo headlights online and on local shops. So, what could we have today? Hmmm…
Since the 2007 Grand Challenge sponsored by DARPA (Defense Advanced Research Projects Agency) is only days far, let us take a closer look at the competitors. For now, I will have my topic focused on the Stanford Racing Team.
“Junior”, a specially modified 2006 Volkswagen Passat, is the Stanford Racing Team’s racer. It is a four-door hatchback powered by a diesel engine and controlled by artificial intelligence (AI). This vehicle will be part of a DARPA research project who goals in developing the ultimate “smart car’ for automated driving. Previously, it has attempted to navigate similar vehicles to run through fixed obstacle courses without any obstructions in motion. However, the upcoming event will be providing new challenge to a whole new level.
David Orenstein from Stanford University Engineering said, “That even means through moving traffic and even obeying California traffic laws.”
Junior is equipped with all kinds of sensors and control input devices mounted all around, which includes a spinning, 360 degree range-finding laser that can provide an accurate topography of the surrounding environment. There are also specially mounted cameras that provide a full 360 degree view. Also, more laser and radar units are found mounted at various points around the vehicles. All were considered necessary to keep the vehicle going until it completes the course.
The data gathered are all being process through a custom developed computer system and logic algorithms. In the process, objects and motion are discerned, which is what is applied to the AI in determining traffic, road and legal conditions around the car. By such, Junior must complete the course with moving traffic together with obeying traffic lights.
Vehicle lead engineer Ganymed Stanek said, “Moving traffic is the big difference to the last grand challenge. Now we need to understand what the other obstacles are doing, taking turns at intersections, need to see who is first at the intersection and drive in the correct order.”
Junior the product of combined efforts of a special team of twelve researchers, scientists, engineers composed of faculty and students of Stanford Artificial Intelligence Lab (SAIL). The School of Engineering provided for the programming and hardware adaptation. The team hopes for a win as they build off their previous 2005 winning entry named “Stanley”, who finished a similar race without moving obstacles or traffic laws.
The team foresees Junior as the answer for those who have physical disabilities that prevent them from driving. Such technology even helps in keeping us safer, as Stanek said, “If the car were smart enough to anticipate accidents, then it could perhaps help us. Of certainly for people who fall asleep at the wheel the car could just take over.”
As of now, the team is still facing their biggest challenge, translating the pixelated computer images and numerical range data into accurate 3D models that can be interpreted and responded to correctly. It wouldn’t be as easy as humans could, so it will have to take a little more time to teach the car. Hopefully, they will get it on time and be one of the 20 finalists at the Urban Challenge Qualifiers held at Victorville by next month.
Among those who fund for Junior are private enterprises, which include Intel, Google, Red Bull, MDV (Mohr Davidow Ventures), NPX, Semiconductors, Applanix, and most especially, Volkswagen.