When defects are missed, turn to technology
It is well known that human inspection is not 100% effective. Fatigue, distractions, and boredom conspire to prevent defects from being spotted, particularly if those defects are not obvious. Detailed examination, potentially using templates and images or taking measurements, is a time-consuming process, and manual inspection is unsuitable for use in a production line with large components. Battery packs manufactured on an automated line require the correct placement of two rubber stickers; these must be completely undamaged and within a predetermined tolerance and must be correctly placed. The surface appearance of the stickers is very variable. The batteries pass under the Inspekto S70 on a conveyor belt at 12 per minute; the reasons for rejection are rubber sticker damage and misplacement.
A dual camera and AI approach makes all the difference
The Inspekto S70 autonomous machine vision system uses a camera and a powerful AI (Artificial Intelligence) image processing system to spot defects whilst minimising false positives. The system is trained using good components and can be configured to ignore product parts that differ by design (such as serial number labels). It is tolerant of orientation and positioning and can scan components placed in the camera’s field of view and on a conveyor belt at rates exceeding 40ppm. These defects are identified in red on the displayed or recorded image to facilitate rework.
The S70
The S70 can be used on many materials/components, including Electrical Components, Metal, PCBs, plastic, and rubber. This capability makes it an ideal solution for products such as injection moulded components where it can identify sink marks, pores, dents and flashing, electrical connectors-presence/missing/position, soldering on PCB’s, integrity & milling swarf in metal components, position and presence of labels, verification of coatings and O ring placement or missing.
