For highest Efficiency and Productivity Beyond Inspection: Optimization of Plastic Film Production

Reliable recognition of defects using SMASH in-line inspecti
Every year, customers demand more and more from an inspection system. It was not too long ago that it was enough for a manufacturer of films and foil to know how many defects occur on the roll, to be able to differentiate between light and dark defects, and to know the size of the defect in order to be able to have enough information about the cause of the problem. Today, customers demand even more precise defect detection data.

This requires crystal clear images, a classification system, which describes the type of defect with the ultimate goal: to detect and put a stop to production glitches at their source. The advantages: The output of a line can be increased while simultaneously raising the level of quality. Customer complaints and the costs associated with them and lost orders altogether can be avoided.

Today’s inspection systems must be able to offer much more than simply the ability to detect defects. In order to increase productivity and the quality of production, accurate defect classification is simply a must for cause analysis as is the ability to generate trend data for process optimization.

What are the overall benefits for the manufacturer? Based on trend figures, for example, systems can clearly indicate at what point the line needs to be cleaned instead of subjecting it to periodic cleaning that may be scheduled much too early. Repeat defects allow insight into the condition of the rolls detailed off-line analyses can help find answers in the event of customer complaints. The ability to accurately classify defects, for example, a nozzle malfunction, makes it possible to quickly take corrective measures. Improved utilization of the equipment, optimized maintenance, and lower operating costs are the direct result.

How does that work? The user is able to define the number of defect classes according to his specifications and customer requests. The QUICKTEACH software, with the ability to teach itself, is responsible for the high defect classification data in real-time. Modern algorithms train the pre-configured system and classify defects, during an ongoing inspection process in a short period of time.. Just a few defect images are all the system needs for the training process. Once a defect is detected, the software compares defect features with the classes it has been trained to recognize and defines the defects based on their corresponding class. At any time during production, new defect classes can be defined. This way, the surface inspection system can reach its maximum performance level, and do it extremely quickly. The user does not need to have any in-depth knowledge of these processes, since the system is easy to use, is neatly arranged and is operated intuitively.

Reliable defect classification is the basis for analyzing the source of defects. Today, users need reliable and extensive information on surface defects along the entire web surface. More importantly, the inspection systems are the tools which provide statistical data which can be processed; this information will allow them to optimize their processes. All of the obtained information is stored in a central database. This information can be combined with other production data. In addition, a great array of trend and statistical analysis is available as well as reports on defects, which can be used for optimizing processes. This will allow all users quick access to concrete results. Production performance can be increased noticeably with these immediate feedback options and wait times in the production process are minimized.

As a significant tool for improving process efficiency, Data Mining complements the SMASH inspection system. The tool gathers the production and defect data and makes it possible to transfer these into any database of choice. With the data provided, the user can perform an analysis by date or time, batch, type of damage, product type or shift and as a result can receive valuable feedback for optimizing production at different processing stages. The advantage for the user is quick, quantified and detailed information.

In addition, using the Enterprise Data Mining tool, statistical data of several systems can even be integrated at multiple production locations. This way, separate production batches located in different parts of the world can be labeled with one standard quality passport.

Not only do state-of-the-art inspection systems detect defects 100% of the time, but they also play a major role in improving process efficiency. They should be among the manufacturer's top choices allowing companies to fully meet the increased demand for quality and production speeds.