Integrated on new PVS units for preforms, this system is setting the global benchmark for visual process control in Rigid Packaging. Key innovations include an interface that uses AI to associate faults with ‘families’ of defects, thus streamlining data interpretation and helping the operator take action
Classy AI is a new SACMI-developed tool that simplifies yet revolutionizes visual process control in the Rigid Packaging field. That revolution begins with making these systems pivotal to modern production processes in order to monitor quality and efficiency and guide direct corrective/maintenance action on the lines.
Classy AI can be applied on the SACMI vision system range, the market standard-setter for specificity, accuracy, reliability and precision. This latest tool doesn’t improve the already state-of-the-art control but, rather, changes the way the system is managed.
With Classy AI, in fact, the operator has a user-friendly dashboard that allows immediate classification of defect types, from the most common to those that even an expert might have trouble interpreting.
So how does it work? Alongside the technical interface featured on all SACMI vision systems, Classy AI provides an innovative dashboard. The left-hand side of the screen shows the detectable defects, grouped into ‘families’ (accidental, technological, decoration-OCR, measures, specific). In the middle, an image highlights the area of the object where the defect has occurred. On the right, instead, analysis results are logged in the form of quality statistics so users can immediately identify any process drifts.
The main advantage of the system lies in being able to recognize not just individual defects but also their origins and where they belong among the various detectable ‘families’. For example, with Classy AI impurities in a preform body can reliably be linked to the problem that caused them (distinguishing, for example, between the classic black spot, which stems from degradation of the plastic, to a defect caused by an insect dropping into the mix).
The operator - not always or necessarily an experienced vision system technician - continues to play a key role in this advanced system: in the event of an anomaly, a specific warning icon appears on the dashboard. This means that even inexpert operators are perfectly able to identify the type of problem, take direct action to solve it or, in any case, filter the alarms that might require the intervention of a specialized technician.
The underlying technology is impressive to say the least: the algorithms needed to run the classification process took the SACMI Rigid Packaging BU Computer Vision R&D team months to develop. The result: a unique solution, quite unlike any other supplied by the world's leading manufacturers of industrial vision systems.
The recent preview showcased the integration of Classy AI on the new PVS 156 system for intensive quality control of the preforms made on the new IPS 300 press. Much more ambitious is the next goal: on the one hand, the aim is to extend the proposal to the full range of SACMI vision systems for Rigid Packaging (caps, preforms, containers, labels and metal). On the other, SACMI aims to take advantage of this new AI application - which simplifies control of complex systems - to ensure such systems become more commonplace on production lines and more effective, with cutting-edge reporting packages leading to better, even automated, feedback control.
Waitin’ 4 you at K 2022! (Hall 13 – 13A63)