Interactive design and development of an intelligent vision-driven 3D printed precision sorting mechanism for silk cocoons

The design and implementation of a small-scale automated cocoon sorting system that utilizes a Raspberry Pi 3 B, a Raspberry Pi 5 MP camera, and machine learning algorithms trained on the Teachable Machine platform are presented in this research paper. Various hardware components were implemented in the system, including a 360-degree servo motor for controlling a rotating disk to feed cocoons, a 100 RPM DC motor for conveyor belt control, and an MG90 180-degree servo motor for final sorting. CAD models were designed using CATIA-V5 and 3D printed with PLA (Polylactic Acid) material to construct the prototype. The system operates with continuous monitoring by the Pi camera, which detects good and bad cocoons, utilizing a trained neural network model to classify the cocoons. Based on this classification, the sorted cocoons are directed to the appropriate destination by the 180-degree servo motor. By automating the cocoon sorting process, improvements in efficiency and accuracy in cocoon classification have been achieved.