MONITORING QUEUES WITHOUT OBSERVING THEM: A NEW TECHNOLOGY FOR INDUSTRY

THIS RESEARCH PROPOSAL FOCUSES ON A NEW DATA-INTENSIVE PROCEDURE DEVELOPED BY DR. RICHARD C. LARSON AT M.I.T., "DEDUCING QUEUE STATISTICS FROM TRANSACTIONAL DATA". TRANSACTIONAL DATA CONSTITUTE THE TIMES OF SERVICE INTITIATION AND SERVICE COMPLETION FOR EACH CUSTOMER SERVED IN A QUEUEING SYSTEM. BY ANALYZING THESE DATA USING RECURSIVE FUNCTIONS, ONE CAN OBTAIN ALL OF THE KEY QUEUEING PERFORMANCE MEASURES FOR POISSON ARRIVAL QUEUEING SYSTEMS. THE NEW METHODOLOGY HAS POTENTIALLY FAR REACHING APPLICATIONS IN BANKING, COMMUNICATIONS, TRANSPORTATION AND GENERAL SERVICE SYSTEMS. THE PROPOSAL AIMS TO DISCOVER THE "NOISE" CHARACTERISTICS OF TRANSACTIONAL DATA AS MEASURED BYPRESENTLY AVAILABLE TECHNOLOGIES, TO ADAPT (IF NECESSARY) DR. LARSON'S METHOD TO REFLECT NOISE CHARACTERISTICS, TO TEST AND EVALUATE USING MONTE CARLO SIMULATIONS THE METHOD IN AT LEAST FIVE APPLICATIONS CONTEXTS, AND TO IDENTIFY ONE OR TWO APPLICATIONS AREAS FOR FURTHER (PHASE II) RESEARCH.