Often customers will purchase a machine vision system in reaction to a customer claim without putting any thought towards the goals they want to achieve. This article highlights the different goals achieved from installing a vision system.
Before purchasing a system, think hard on how to use an inspection system. What are the goals?
- Stop embarrassing defects getting to your customer.
- Increase productivity.
- Reduction in substrate waste.
- Fault identification so as to enable process improvements.
Stop embarrassing defects getting to your customer.
In the section prior to this we discuss the output system of an inspection system. Alarms and display systems are good to alert operators to problems. This will allow early intervention and will work well to reduce waste. However when running an operation with roll to roll converting, it can be difficult to remove waste.
Marking a defect is not always an option. If defects are common, or when they occur go on for some time, defect marking systems can cause a lot of problems. It may be that the end result will be what we call a hedgehog. A roll with hundreds of tags sticking out. Marking systems are more expensive and if the web moves also require an edge guide. What is required is a smart way to use the data generated by the system while synchronizing the data to an event on the web. Typically it can be possible to use a splice or a large defect, or a small number of tags as zero events to allow waste removal.
So investigate what data is stored, the data format, and if it is possible to reverse the data to match the next process and to use zero out events for synchronization.
How to use the data to increase productivity. What is the goal here? To run faster while using less resources. Inspection at least by humans can be a slow and repetitive process. Inspection systems should allow operators increase line speeds, as the inspection system if specified will not care if you are running 20% faster.
What is required is to understand what tools are available to allow quality departments check production runs to see if there is any requirement for rework prior to converting, or in some cases when automated waste control systems are used, prepare an automated converting strategy for the next process. This is often referred to as "Smart converting". Ask your supplier if this is part of their offering.
Reduction in substrate waste
The key to reducing waste is to identify problems prior to each process. If the converter is producing their own substrates, then adding inspection stations at each process will pay back very quickly if substrates represent a large % of the cost of operations. Examples of this is to inspect plastic film and foil before used in the production of flexible packaging.
By analyzing the quality of the film, it may be possible to use the poorer grades during the make ready for the next conversion process. The net result is that you will be using waste from the prior process during the high risk start up time of the second. Rather than adding to waste you are optimizing your chance to produce higher yields when the second process is running well.
What is required is the ability of your quality system to use data from a variety of inspection solutions to optimize quality.
Fault identification so as to enable process improvements
Inspection systems produce a large amount of data. This data should as described above be used in feed forward operations to enable smart converting. By integrating the output of the inspection system with a historian, the data can be used for regression analysis and fault identification.
The key features and capability to seek are:
- Ability to integrate with historians
- Ability to stream video to disk to enable process improvements
- Open formats to allow 3rd party tools to analyze the data.
Do not focus on detection alone when specifying an inspection system. Consider how the equipment will be used as described above and spend the time to understand the offerings from all vendors. An inspection system can pay back in months not years if the data is used correctly.
Download our latest whitepaper on specifying your vsurface inspection system by following the link below.