Smart Robot for Collection and Segregation of Garbage

Abstract

This paper proposes a robotic bot to collect the litter thrown and segregate the garbage so as to reduce human effort and keep the environment clean. The proposed robotic bot autonomously identifies garbage and classifies it into biodegradable and non-biodegradable. The final collected garbage is, in a way, segregated automatically. The robot moves towards the garbage just like a rag picker does. When the robot's bins are full, a signalling system is created. This autonomous bot is capable of collecting, segregating and providing signals if the garbage bin is full. The yolov4 model was used to train a dataset of 10,000 images of both the categories and the model was able to detectthe test images with adequate efficiency of 99.66 percent, making it promising candidate for using as an alternative to human effort.