Having won the 2005 Grand Challenge in the Nevada desert with their autonomous car ‘Stanley’, Stanford Racing Team (SRT) passed a milestone in artificial intelligence (AI). Now, the 2007 Urban Challenge presents a major new element that will push technology into new territory. Not only will the robot vehicles have to negotiate stationery obstacles, they have to contend with moving traffic as well. Effectively they will need to understand their environment, instead of just sensing it.
The 2007 Urban Challenge therefore presents an excellent opportunity to develop innovative approaches to autonomous guidance systems and machine vision. This is why NXP, at the forefront of vehicle electronic systems, is supporting the team’s effort. Although this project is in its early days, we anticipate the lessons learned during this event will contribute to ‘real-car’ systems in the very near future.
Introducing ‘Junior’
‘Junior’ – named after Stanford University’s founder Leland Stanford Jr. – is a 2006 Passat with a few differences. The steering, throttle and brakes all have been modified to be completely computer-controllable. Engineers have also created custom mountings for a suite of sophisticated sensors that allow ‘Junior’ to interpret and react in real-time to the vehicle’s actual surroundings.
Stanford Racing Team
Stanford Racing Team is based in the School of Engineering at Stanford University and is supported by returning industry team members Intel, MDV-Mohr Davidow Ventures, Red Bull and Volkswagen of America. This year they are joined by new supporters Applanix, Google and NXP Semiconductors.
DARPA Urban Challenge
The Defense Advanced Research Projects Agency (DARPA) will hold its third competition for autonomous vehicles on November 3, 2007. This year’s Urban Challenge adds a new dimension for entrants. The competing vehicles will have to complete missions in a simulated city environment, including motion of the other robots and traffic laws. This means that on race day the robots will have to master concepts such as right of way – which includes a component of prediction. Predicting future actions is a tricky problem, but is crucial for safe driving.