ISO-Spector M1A from Marantz Electronics (MEK) delivers a self-learning algorithm for solder joint inspection
One of the greatest challenges in programming AOI, (Automatic Optical Inspection) systems is the detection of all non-conforming placements, especially those related to solder joints, without creating a long list of False Calls. The industry has always faced this problem and has had to accept that the more accurately you wish to measure, the more false judgements a system will make. It commonly results in long debugging time and complex programming.
To overcome this challenge, the full 3D ISO-Spector M1A from Marantz Electronics (MEK) delivers a self-learning algorithm for solder joint inspection that will detect any deviation outside the expected standard appearance of a solder joint. Artificial Intelligence is continuously and centrally monitoring production and adjusting hundreds of tolerance values where needed to maximise detection and minimise false calls.
The programmer does not have to specify the specific inspection locations, light settings or acceptance criteria. This not only reduces programming time, but more importantly removes the often-unpredictable human variable to ensure that the inspection results are more reliable.