APPLICATION OF ARTIFICIAL INTELLIGENCE TO AUTOMATED RECYCLING

DIVERSION OF PLASTICS FROM LANDFILL TO RECYCLING CAN POTENTIALLY SAVE THE ENERGY EQUIVALENT OF 60 MILLION BARRELS OF OIL ANNUALLY AND REDUCE LANDFILL VOLUME REQUIRMENTS BY UP TO 20%. THE U.S. ENVIRONMENTAL PROTECTION AGENCY HAS RECOMMENDED THAT SOURCE REDUCTION, RECYCLING, VOLUME REDUCTION, AND LANDFILLING BE APPLIED, IN THAT ORDER, IN THE TREATMENT OF MUNICIPAL SOLID WASTE (MSW). RECYCLING HAS RECENTLY BECOME A MAJOR COMPONENT OF MUNICIPAL WASTE MANAGEMENT PROGRAMS. HIGH COST OF LABOR INTENSIVE CURBSIDE RECYCLING PROGRAMS HAVE BROUGHT ABOUT A SECOND GENERATION TECHNOLOGY, MIXED WASTE PROCESSING, FOR ACCOMPLISHING HIGHER LEVELS OF RECYCLING AT REDUCED COSTS. THE ECONOMICS OF MIXED WASTE PROCESSING DEPENDS UPON EFFICIENT AUTOMATED PROCESSES FOR RECOVERING RECYCLABLES FROM MSW. EFFICIENT AUTOMATED TECHNOLOGY EXISTS FOR RECOVERY OF STEEL, ALUMINUM, COMPOSTABLE FOOD WASTE AND PAPER PRODUCTS FROM MSW. THESE ITEMS MAKE UP THE BULK OF MSW. HOWEVER, THE ONLY EXISTING METHOD TO RECOVER GLASS AND PLASTIC CONTAINERS FROM MSW IS TO MANUALLY HANDPICK THEM FROM THE WASTE. THE OBJECTIVE OF THE PHASE I RESEARCH PROGRAM IS TO DETERMINE FEASIBILITY FOR DEVELOPMENT OF AN EFFICIENT AUTOMATED PROCESS FOR RECOVERING POST-CONSUMER PLASTIC AND GLASS CONTAINERS FROM MSW USING RECOGNITION ALGORITHMS INCORPORATING ARTIFICIAL INTELLIGENCE TECHNIQUES.