Artificial Intelligence Hybrid Range Scheduler

Scheduling is a complex, NP-hard problem that challenges the use of any single solution technique. In Phase I, we built a model that uses our multi-technique Quantum Leap Problem Solving Engine to solve a representative but simplified version of the AirForce Satellite Control Network (AFSCN) contact scheduling problem. The system demonstrated that it could schedule the hundreds (500 to 600) of tasks per day required well within the expected time frame of a few hours for next-day scheduling. We arepresently creating a scheduling language and investigating new techniques for improving performance.In Phase II, we will create a Scheduling and Planning Framework (SPF) which builds upon the phase I proof of concept and incorporates the new performance improving techniques. We will build a scheduling system for the AFSCN problem that goes beyond thesimple proof of principle and solves the real scheduling problem. We will also solve the spacelift range asset scheduling problem. Both solutions will be built on the SPF using unique data input and output interfaces.