NETWORK OF LEARNING SYSTEMS FOR AIR TRAFFIC PLANNING AND CONTROL A PILOT STUDY WILL INVESTIGATE HOW TO PLAN, SCHEDULE & CON-

CONTROL AIR TRAFFIC UTILIZING A NETWORK OF EXPERT LEARNING SUBSYSTEMS. THE SYSTEM, WITH SIMULATIONS, WILL DEMONSTRATE PROBLEM SOLUTIONS AND PROVIDE DIRECTION FOR FUTURE RESEARCH EVALUATION, IMPLEMENTATIONS, AND COST. THE R & D IS GUIDED BY DR. ED PATRICK'S 28 YEARS OF PIONEER-ING WORK IN PATTERN RECOGNITION, ARTIFICIAL INTELLIGENCE, ANPRODUCT R & D OF LEARNING SYSTEMS MORE GENERAL THAN CONVEN- TIONAL THREE LAYER NEURO NETS, AND BY HIS EXPERIENCE AS AN INSTRUMENTAL RATED, LICENSED PILOT. SOME MODULES IN THE NETWORK WILL UTILIZE TRAINING EXAMPLES GENERATED BY COMPUTER SIMULATION (WITH PROBLEM CONSTRAINTS) AND EXPERT KNOWLEDGE; AND OTHERS WILL USE PHYSICAL EQUATIONSFOR AIR TRAFFIC PLANNING. OUTCOMES CAN BE PREDICTED FOR FINDINGS NEVER PREVIOUSLY SEEN.