The Defense Advanced Research Projects Agency, a government research arm, invited 35 teams to compete in a driverless vehicle race. One of the expected participants is the Team Caltech, a robotics group at the California Institute of Technology.
The Team Caltech is in the process of fine-tuning - not typically just putting in premium Saab tune up parts or the-likes but through using more sophisticated measures. They have been doing series of tests and trials with a large self-driving van, a Ford E-350 weighing 8,000 pounds with armor plating. It is in preparation for this month’s semifinals of the Urban Grand Challenge.
One of the rules in the race says that if a robot hits another vehicle, its team is automatically disqualified from the competition. Such case is highly possible to those newbie machine drivers. In the case of the victim of a collision, its team is given 30 minutes to fix their machine and continue in the race. However, if the damage is too much, the team will be out of the competition.
The Team Caltech seems to be highly confident that they can survive until the end of the challenge, surpassing any crash that their machine might get into. Joel Burdick, professor of mechanical engineering at the California Institute of Technology, and Team Caltech’s co-leader, said, “We’re one of the biggest robots in the race. We think we’d be a physical victor in any crash.”
The challenge does not only require the robots to have physical hardware to win the competition. The teams must be able to create software that can synthesize and interpret the data coming in from multiple external sensors on the car, including radar, ladar, cameras and GPS. Such software should also have the ability to plot a drivable path on city roads, avoid real-time obstacles and follow traffic laws.
If people have a hard time doing such mission, how much more for the robots? But, it isn’t impossible, since the teams’ eagerness is boosted by a $2 million first place price. The difficulty is worth that much. So, each team prepares very well for the event on October 26, in Victorville, California, at the site of the former George Air Force Base.
DARPA offered $1 million to 11 track “A” teams, including the Team Caltech. It is their way to support the top participants and competitive challengers in the Urban Grand Challenge. That’s one of the reasons why Caltech currently have enough money to prepare.
In 2005, Caltech build its robot, “Alice”, spending approximately $200,000 in the process including travel expenses. Caltech will be using 15 percent of the original software used in their robot. And, the rest are new codes. The software is comprised with three parts: a mission planner that takes in data from DARPA to sequence that course of the race, a traffic planner that gets data from sensing technology on factors like the speed of other cars on the road; and a trajectory planner to execute on the above data. There are still small errors in the program, but Caltech is already working on it.
The dedicated Caltech team is composed of about seven full-time graduate students, one post-doctoral student and two faculty members. The members are from several departments, of which include computer science and mechanical engineering departments. They are supported by Caltech’s Jet Propulsion Lab, NASA’s research center, with regards to the computer vision system. A professor in Caltech’s control and dynamical systems department, specializing control theory, Richard Murray is held as the team’s three-time leader. And, Burdick sees the expertise in control theory, a branch of engineering and mathematics that studies the behavior of dynamic systems as an advantage for the team to stand out in the competition. He said, “This is a large-scale computing problem. We’re going to have some unique, more structured approaches to planning algorithms.”
Well, that would mean we could expect a great performance of the Team Caltech in the DARPA Urban Challenge. And, they are one of the closest possible qualifiers in the finals on November 3.
