THE DEVELOPMENT OF NON-ORGANIC WASTE WEIGHT CALCULATION SYSTEM UTILIZING MACHINE LEARNING

• Main Author: Wie Jonathan, Jason
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/77342
• Institution: Institut Teknologi Bandung

Waste is a problem present all over the world, including in Indonesia. Indonesia is among the countries that produce the highest amount of non-organic waste in the world, especially plastic waste. Currently, waste calculation at temporary shelter (TPS) is still carried out manually by sanitation workers. The amount of waste in a TPS is considerable, and it takes a significant amount of time to complete the waste calculation process. Therefore, this research objective is to develop an alternative solution to address this waste calculation issue, which is a waste weight calculation system utilizing machine learning. This system employs machine learning to recognize provided waste images. The types of waste that can be recognized by the system are non-organic waste, particularly beverage packaging waste. Once the waste is identified, the waste calculation process will be carried out automatically. The proposed system will be implemented in a mobile application. The method employed in this research is the waterfall method, consisting of several stages: requirements analysis, design planning, coding, testing, and deployment. For the testing phase of the developed system, various tests were conducted, including functional and non-functional requirement testing, system accuracy testing, and waste calculation testing. The results show that the system has successfully operated in accordance with predefined functional and non-functional requirements. Furthermore, for the system accuracy testing, a result of 78% was achieved, which is considered good. Lastly, for the waste calculation testing, the system was able to perform automatic calculations for non-organic waste types.