Kat-5: systems based on a successful paradigm for the development of autonomous ground vehicles

We have developed several systems that were combined to create “Kat-5,” one of four successful competitors in the 2005 DARPA Grand Challenge. Although each of these systems has its individual merits, they are all products of a single paradigm that has served the vehicle well during its development and testing. This strategy can be stated succinctly: simple solutions to simple problems and complex solutions to complex problems. We explore the navigation systems designed under this strategy and their resulting performance, both during initial testing and during the actual Grand Challenge Event. The systems discussed include obstacle avoidance, obstacle detection, and geospatial data storage and processing. The results of the implementation of these systems have been exciting, often exhibiting behavior that was previously thought to require much more complex systems. Kat-5’s success at the 2005 DARPA Grand Challenge can be directly attributed to the systems presented here.

[1]  Cris Koutsougeras,et al.  KAT-5: Robust systems for autonomous vehicle navigation in challenging and unknown terrain , 2006, J. Field Robotics.

[2]  Håkan Jonsson,et al.  An obstacle-avoiding minimum variation B-spline problem , 2003, 2003 International Conference on Geometric Modeling and Graphics, 2003. Proceedings.

[3]  A. A. Ball,et al.  An investigation of curvature variations over recursively generated B-spline surfaces , 1990, TOGS.

[4]  Thierry Fraichard,et al.  From Reeds and Shepp's to continuous-curvature paths , 1999, IEEE Transactions on Robotics.

[5]  Carl de Boor,et al.  A Practical Guide to Splines , 1978, Applied Mathematical Sciences.

[6]  Renato Pajarola,et al.  LOD-based clustering techniques for efficient large-scale terrain storage and visualization , 2003, IS&T/SPIE Electronic Imaging.