Desarrollo de un sistema automatizado de detección y exclusión de alcancías erróneamente posicionadas a base de visión artificial y extracción por pistón

Abstract

The main objective of this thesis is to create a system that optimizes efficiency in the processes at ONTUBE S.A. by identifying and excluding misaligned tubes. This device is based on the development of a vision system trained to identify incorrectly positioned tubes on a production line. By implementing advanced vision technologies, the goal is to enhance quality and effectiveness in production, reducing errors and raw material losses. For this device, a vision system integrated into a Raspberry Pi will be used, which will send a signal to a solenoid valve when it detects a misaligned tube. This signal will activate compressed air, which in turn will activate a double-acting piston to exclude the defective tube from the production line. Finally, weekly loss data were analyzed to compare the device's efficiency. The results demonstrate the feasibility of the proposed system, showing a significant improvement in the production line's efficiency.