Plan for 2009-2010
We need to reconstruct this information since the wiki crashed and hadn't been backed up. Please start by listing what technologies you are working on long-term, and what you are working on for the November field trial.
Spencer Gardner
Spencer Clark
Long-term I think I'm working on my GPS-related code and on Phairwell. For the field trial, I am working on some nice at-a-glance widgets for Phairwell. When/if I am finished with that, I hope to make some other changes to Phairwell.
Lanny Lin
Extending terrain feature-based modeling of lost person behavior. Specifically, adding trail following and intended destination to the model, and also using real GPS track logs and possibly real WiSAR scenarios for validation.
Integrate distribution generation with SkyDriver as a client/server component.
Expand the distribution generation component with interactive interface that allows the user to modify distribution dynamically.
Improve path-planning algorithms with coarse-to-fine and gradient descent on the Global Warming dimension to speed things up.
Integrate path-planning with SkyDriver as a client/server component.
Investigate how partial detection and seeability affect algorithms' performance.
(Possibly) adding time dimension to the path-planning problem.
Implement sliding autonomy on SkyDriver and perform user study on simulated environment. Also evaluate sliding autonomy with real field trial.
For November field trail, I better not integrate any of these to reduce risk.
Back to top