|
| ||
course participants course announcements about this wiki questionnaires and assignments slides of presentations course schedule related resources Gerhard Fischer Hal Eden Mohammad Al-Mutawa Ashok Basawapatna Lee Becker Jinho Daniel Choi Guy Cobb Holger Dick Nwanua Elumeze Soumya Ghosh Rhonda Hoenigman elided#1 Dan Knights Kyu Han Koh elided#2 Yu-Li Liang Paul David Marshall Keith Maull Jane Kathryn Meyers John Michalakes Michael Wilson Otte Deleted Page Joel Pfeiffer Caleb Timothy Phillips Dola Saha deleted  |
We present a new machine learning based approach for robotic path planning in unknown unstructured outdoor environments. The proposed system is tested in simulation and also on a variety of real robotic platforms, and compared to other state-of-the-art systems. In order to demonstrate that out technique is both viable and superior to previously proposed techniques, we evaluate the proposed technique in a large number of heterogeneous environments. We also demonstrate that the proposed system is not vulnerable to common pathological cases that plague other implementations, by testing our system on contrived courses that are specifically created to invoke such failures. We also present a discussion of our system's weaknesses and outline the circumstances under which it should not be used. Over all, we find that our system is robust and generally outperforms other similar systems in a wide variety of unknown outdoor unstructured outdoor environments. Last modified 12 December 2007 at 9:09 am by MOtte |