This article describes the robot Stanley, which won the 2005 DARPA Grand Challenge. Stanley was developed for high‐speed desert driving without manual intervention. The robot's software system relied predominately on state‐of‐the‐art artificial intelligence technologies, such as machine learning and probabilistic reasoning. This paper describes the major components of this architecture, and discusses the results of the Grand Challenge race. © 2006 Wiley Periodicals, Inc.

Summary. This article describes the robot Stanley, which won the 2005 DARPA Grand Challenge. Stanley was developed for high-speed desert driving without human intervention. The robot's software system relied predominately on state-of-the-art AI technologies, such as machine learning and probabilistic reasoning. This article describes the major components of this architecture, and discusses the results of the Grand Challenge race.

Abstract-We present a method for identifying drivable surfaces in difficult unpaved and offroad terrain conditions as encountered in the DARPA Grand Challenge robot race. Instead of relying on a static, pre-computed road appearance model, this method adjusts its model to changing environments. It achieves robustness by combining sensor information from a laser range finder, a pose estimation system and a color camera. Using the first two modalities, the system first identifies a nearby patch of drivable surface. Computer Vision then takes this patch and uses it to construct appearance models to find drivable surface outward into the far range. This information is put into a drivability map for the vehicle path planner. In addition to evaluating the method's performance using a scoring framework run on real-world data, the system was entered, and won, the 2005 DARPA Grand Challenge. Post-race log-file analysis proved that without the Computer Vision algorithm, the vehicle would not have driven fast enough to win.

Quenched QCD simulations on three volumes, 8 3 ×, 12 3 × and 16 3 × 32 and three couplings, β = 5.7, 5.85 and 6.0 using domain wall fermions provide a consistent picture of quenched QCD. We demonstrate that the small induced effects of chiral symmetry breaking inherent in this formulation can be described by a residual mass (m res ) whose size decreases as the separation between the domain walls (L s ) is increased. However, at stronger couplings much larger values of L s are required to achieve a given physical value of m res . For β = 6.0 and L s = 16, we find m res /m s = 0.033(3), while for β = 5.7, and L s = 48, m res /m s = 0.074(5), where m s is the strange quark mass. These values are significantly smaller than those obtained from a more naive determination in our earlier studies. Important effects of topological near zero modes which should afflict an accurate quenched calculation are easily visible in both the chiral condensate and the pion propagator. These effects can be controlled by working at an appropriately large volume. A non-linear behavior of m 2 π in the limit of small quark mass suggests the presence of additional infrared subtlety in the quenched approximation. Good scaling is seen both in masses and in f π over our entire range, with inverse lattice spacing varying between 1 and 2 GeV.

No abstract