Timeline

I hope to complete this thesis in a little over a year. Here is my target schedule:

Jan 96

Basic research: ROUT (done)

Mar 96

Basic research: optimization as MDP; STAGE (done)

May 96

Develop initial codebase with applications: VLSI channel routing; 2-D radiotherapy treatment planning (done)

Jun 96

Propose

Jul 96

Application focus: factory scheduling

Aug 96

Work on algorithm refinements

Sep 96

Application focus: 3-D radiotherapy treatment planning

Oct 96

Application focus: spacecraft configuration (at JPL)

Nov 96

Finalize algorithms

Dec 96

Write release-quality software

Feb 97

Gather experimental results

May 97

Write dissertation

Aug 97

Defend

This thesis and timeline face several risks. First, real applications inevitably introduce thorny issues and software complexity. Fortunately, the simulation software for all of the major domains but one (NASA) is already implemented, and the NASA implementation is to be done by a colleague at JPL. Second, some of the algorithms I have proposed to investigate are quite CPU-intensive. Thoroughly evaluating the cross-product of algorithms domains is out of the question; I will have to judiciously choose the subset of experiments that answers the most interesting questions.