BINSIGHT: Waste Sorting Enhanced with Machine Learning and Reward System

The researchers improved waste segregation procedures at Pamantasan ng Lungsod ng Maynila (PLM) by creating BINSIGHT, an inventive IoT-based waste sorting system that uses machine learning and reward system. The waste classification system uses FOMO (Faster Objects, More Objects) MobileNetV2 0.35 and is based on the Raspberry Pi platform. Performance testing showed outstanding classification capabilities with an overall F1 score of 94.7%. Paper waste achieved an accuracy of 80%, while food packaging and plastic waste had ideal accuracy rates of 100%. By automation and incentives, BINSIGHT seeks to improve waste management procedures while encouraging friendly disposal methods among campus residents.The system underwent comprehensive evaluation using ISO 25010:2011 criteria, and 52 PLM staff and students took part. The findings showed outstanding performance on all important metrics: 98% of respondents agreed that it was functionally appropriate, 90.4% affirmed that it worked efficiently, 92.3% said that it operated without errors, and 98% accepted that the security measures were strong. Achieving an overall mean score of 3.71 across evaluation criteria, BINSIGHT demonstrates significant potential to overcome waste management challenges. These results show that the system successfully encourages proper recycling behaviors and supports environmental sustainability initiatives, providing an achievable solution for enhanced waste management at PLM.