Full Stack Developer

In my last project, we created a system to prevent work accidents in a foundry industry. The process involves evaluating molds on the conveyor belt before receiving liquid iron (+/- 1350ºC) for the production of parts. These molds must be in perfect condition; otherwise, they need to be removed. Originally, a team did this inspection exposing themselves to accident risks. To solve this problem, we created a web system (Laravel, Vue, Node, Bootstrap, React, Postgresql) along with a robotic arm (C language, Raspberry Pi, camera, motors and various sensors). In this way, we capture images of the molds and analyze them using a Convolutional Neural Network (CNN) and deep learning trained with more than 3000 real and synthetic data images, achieving an accuracy rate currently of 95%. A BackOffice team monitors the process through the web system that shares the operation data and also the control of the robotic arm, providing emergency or preventive actions with the collaborators present on site. Among all the projects I participated in, this was undoubtedly the most comprehensive and required the greatest variety of knowledge, covering fields such as calculus, trigonometry, electronics, deep learning, frontend, backend, embedded system, edge computing and 3d modeling. The project is currently undergoing changes and evaluation to be scalable.