Quality control is an important step for every production line. For a long time, it was depending on the eyes, hands, and experience of experts, but with the rise of sensors, signal processing and AI new approaches became available. Today controllers such as specialized microscopes can use devices that often include multiple different types of sensors to analyze even the smallest parts of a product. To make these devices faster, more efficient and to increase their level of automatization, the integration of AI into or near them is an important next step. To contribute to this development, IFAG explores, develops, and evaluates a tinyML ANN ASIC and corresponding algorithms for color classification at the edge, since color sensor are an important part in many quality control processes.
