Typically human complex inspections are a challenge to be won with deep-learning e vision in sight. We know that the collaborative robotics is, to date, an efficient and functional solution to project us towards the future e help companies out of the crisis. However, a question may arise: it is possible to try to replace and automate typically human actions such as complex inspection operations in sectors such as automotive, consumer electronics, consumer products, packaging, food & beverage , healthcare and logistics?
Yes. In this area too they are emerging specially designed platforms, intended for example for optical character recognition (OCR), assembly verification, and the detection of product defects that are difficult to identify with traditional industrial vision tools. These are vision solutions enriched by deep learning and therefore capable of solving inspections carried out by humans would be excessively complicated, time-consuming or expensive to program using traditional computer vision.
With the incorporation of deep learning software on in-sight vision devices, even the most complex visual inspections can be solved. Devices of this type have recently been presented on the Italian market and combine high self-learning capacity with a vision system that even expands the limits of what can be inspected in industrial automation. At the base is a system that can be “trained” using sample images in order to be configured.