I will proceed in using the forum as a document for my design, changes, errors and needs regarding the Ant Challenge.  

The C++ starter package while usable needs some serious design overhaul for robust AI algorithms.   I have been able to implement breadth-first-search(BFS) for ants to seek food.  While this is a critical step, it is not the only one needed.  In observing the trials based on food seeking, I notice the ant have no ability to attack a hill (the search for food may have brought them close to an enemy base), how to protect the base (again the search for food may drive them away from the base), and how to interact with enemy ants in sight.

This leads to the need for a functioning orders distribution and decision system.  The implementation of this architecture will make the process of dynamic decision changing more easily performed as opposed to performing BFS for food and then deciding going for food is a bad choice given the current conditions.  Additionally this leads directly to separating BFS for food, BFS for Exploration, and BFS for defense and attacking.

I am currently exploring the implementation of the order architecture via the Boost Bimap Library.  I believe this to be comparable the the Gauva bimap data structure.  I am currently have an issue comprehending the state my class variables need to be in to accomplish the integration of the Bimap data structure.  I will work through this over the next several days and update the forum.  Feel free to leave comments or questions.