Autonomous SAR e-ATV
The ISL websites for this project are here (for the Hallway Hummer) and here (for the Minimoto eATV). Logs are also maintained by the current ISL AIBO Research Team:
Logs
Looked up information regarding the DARPA 2005 Grand Challenge and any vehicle entries with similarities to our e-ATV. Got the following papers from the Journal of FIeld Robotics, 23(9):- S. Thrun et al., "Stanley: The Robot that Won the DARPA Grand Challenge," pp. 661-692.
- W. Travis et al., "SciAutonics-Auburn Engineering's Low-Cost High-Speed ATV for the 2005 DARPA Grand Challenge," pp. 579-597.
- A. Lakhotia et al., "CajunBot: Architecture and Algorithms," pp. 555-578.
The Thrun paper features Stanley - I figured this is a good reference on "winning" ideas for such a challenge (: The Travis paper features Rascal, an ATV, which is whay I chose to download it as a reference. The Lakhotia paper features CajunBot, a 6-wheeled ATV - there may be some good ideas here, too. The e-ATV Team will have to look into these papers for some ideas ...
Algorithmically, it looks like we are going to have to use SLAM - Simultaneous Localization and Mapping, which was originally proposed by Smith, Self, and Cheeseman (1990) in their paper "Estimating Uncertain Spatial Relationships in Robotics." Another good reference is Dissanayake et al.'s (2001) paper "A Solution to the Simultaneous Localization and Map Building (SLAM) Problem" ...
Posted on February 22, 2007
Met with Shane and Nalaka to discuss hardware issues with the Minimoto. Plans include getting rid ot the plastic shell covering the chasis, locating something similar to a roof rack/basket to mount on the chasis, and identifying actuator/motor types for steering, braking, and acceleration. We also looked into the possibility of improving the turn radius of the vehicle by cutting out a stopper (probably installed by the manufacturer for safety) at the bottom of the vehicle.
The roof rack/basket mount facilitates loading of computing and sensory equipment on the Minimoto. Using a basket also allows us to have cables hanging through the basket if necessary. We hope to order some necessary parts, soon, so the Minimoto transformation can begin ...
Posted on February 8, 2007
I've had some ideas for this project over the break and early this Spring 2007 semester. The Team has been working on getting the Hallway Hummer up and running, but for both the Hummer and the Minimoto, I insist that
- we handle uncertainty using fuzzy logic (see entry in Wikipedia); this also allows the control and sensor fusion to be dynamic, just like its anticipated (unknown) environment ...
- we implement a wireless remote monitoring system that keeps track of (and logs) all system functions (e.g. sensor inputs)
Posted on January 29, 2007