Oct 8, 2007
The Victor Tango team volunteered to give a lab tour and presentation to Mechanical Engineering students at Virginia Tech. With all of the of the team members either being full time students or having strong ties to academia, the team highly values the dissemination of information. Team Victor Tango's lab, referred to jokingly as the "bot cave" by team members, is an excellent setting for learning. There are two vehicles fully outfitted for autonomous capabilities, a third vehicle in the process of being converted, and a plethora of sensors and actuators decorating the shelves.
The students were part of an Advanced Instrumentation class taught by Dr. Al Wicks. The main focus of the visit was to discuss the vehicles' sensor suites, detailing the types of transducers, signal processing techniques, and methods of sensor data fusion. There are three major sensors that the vehicle uses: Laser Range Finders (LRFs), cameras, and Global Positioning System (GPS). LRFs, or LIDAR, sensors are used mostly for obstacle detection and road contour detection (curbs and berms). These sensors send out laser beams that reflect off of surfaces and return to the sensor. By knowing the time it takes for the beam to return, the sensor can figure out within millimeters how far away the surface is. The team is taking advantage of advancements in sensor technology made by IBEO, which now provide the ability to scan in 3D and differentiate between cars, people, and static objects such as large debris and curbs. Cameras are used mostly for lane detection, similar to lane departure warning systems currently available in high end vehicles. GPS is the main method used to localize the vehicle. The GPS used by the team is far more accurate than the TomTom sitting on your dashboard. While most commercially available systems have accuracy in the vicinity of 2 meters, the team's Novatel unit can locate the vehicle within 0.1 meters of its actual latitude and longitude. By combining these sensors, in addition to a few others, Odin is able to perceive his surroundings in an incredibly accurate manner- probably better than many of the people you pass in your car on a day to day basis.
Saturday, October 13, 2007
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment