DARPA Urban Challenge: The basics

One of the earliest milestone that an autonomous car needs to reach for the competition is to be able to follow GPS waypoints. Since I don't have too much time to detail Kalman filtering using GPS and an IMU, I will point to the explanation and implementation of a software that does just that for the Texas A&M University autonomous car that was never entered in the competition. Two theses were filed on the subject matter:

One was written by James Massey on Control and way point navigation of an autonomous ground vehicle

Abstract: This thesis describes the initial development of the Texas A&M Autonomous Ground Vehicle test platform and waypoint following software, including the associated controller design. The original goal of the team responsible for the development of the vehicle was to enter the DARPA Grand Challenge in October 2005. A 2004 Ford F150 4x4 pickup was chosen as the vehicle platform and was modified with a 6” suspension lift and 35” tires, as well as a commercial drive-by-wire system. The waypoint following software, the design of which is described in this thesis, is written in C and successfully drives the vehicle on a course defined by GPS waypoints at speeds up to 50 mph. It uses various heuristics to determine desired speeds and headings and uses control feedback to guide the vehicle towards these desired states. A vehicle dynamics simulator was also developed for software testing. Ultimately, this software will accept commands from advanced obstacle avoidance software so that the vehicle can navigate in true off-road terrain.

It is available at: http://handle.tamu.edu/1969.1/3862

The other one was written by Graig Odom entitled Navigation solution for the Texas A&M autonomous ground vehicle

Abstract: The need addressed in this thesis is to provide an Autonomous Ground Vehicle (AGV) with accurate information regarding its position, velocity, and orientation. The system chosen to meet these needs incorporates (1) a differential Global Positioning System, (2) an Inertial Measurement Unit consisting of accelerometers and angular-rate sensors, and (3) a Kalman Filter (KF) to fuse the sensor data. The obstacle avoidance software requires position and orientation to build a global map of obstacles based on the returns of a scanning laser rangefinder. The path control software requires position and velocity. The development of the KF is the major contribution of this thesis. This technology can either be purchased or developed, and, for educational and financial reasons, it was decided to develop instead of purchasing the KF software. This thesis analyzes three different cases of navigation: one-dimensional, two dimensional and three-dimensional (general). Each becomes more complex, and separating them allows a three step progression to reach the general motion solution. Three tests were conducted at the Texas A&M University Riverside campus that demonstrated the accuracy of the solution. Starting from a designated origin, the AGV traveled along the runway and then returned to the same origin within 11 cm along the North axis, 19 cm along the East axis and 8 cm along the Down axis. Also, the vehicle traveled along runway 35R which runs North-South within 0.1°, with the yaw solution consistently within 1° of North or South. The final test was mapping a box onto the origin of the global map, which requires accurate linear and angular position estimates and a correct mapping transformation.

http://handle.tamu.edu/1969.1/4244

I'll mention later how our vehicle differs from this approach.