Speed, Flexibility, Experience Equal Success in Surface Inspection Market

The world has watched as petroleum prices skyrocket past $45 a barrel to $50 and beyond, increasing the cost of most products manufactured around the world either by increasing raw material costs or transportation. Every plastic wrapper starts as a thin film, which starts as petroleum. Other raw material markets, from optical-film chemical components to metals packages for children's drinks, have also increased in costs as natural resources are stretched across a growing global manufacturing landscape.

In the surface inspection market, regardless of whether the surface is a thin-film on an LCD panel, a package for fruit snacks, or a newspaper, machine vision is helping manufacturers improve their productivity by inspecting both incoming raw materials and outgoing products. What was once seen as a technology for savvy manufacturers wanting a competitive advantage has become a business necessity for the web manufacturing industry.

‘‘We no longer have to do pioneering sales efforts to get the technology accepted,’‘ says Jim Borges, CEO of i2S North America (Milford, Connecticut), a supplier of line scan hardware and software. ‘‘We're at the point that the technology is very well accepted.’‘

Two types of surface inspection
Surface, or web, inspection falls into two general categories: pattern and non-pattern inspection. Pattern inspection typically refers to inspecting surfaces that have been altered in some fashion, such as printing on paper, plastic, metal or other material. Non-pattern inspection typically refers to inspecting raw materials, such as paper, plastic films, metals, etc.

For a company to be at its most productive, it needs to know the condition of incoming materials as well as outgoing products to optimize processes. ‘‘Many end-users depend on inspection because they don't want to add value to defective materials,’‘ explains Werner Goeckel, president of ISRA Surface Inspection Americas (Norcross, Georgia). ‘‘With the escalation of raw material costs, like oil, you don't have time for that. And it's not just for high-value products but for low-value too. Increasing yields and really understanding their processes to understand the cost and effect of defects is pushing the [surface inspection] industry. Surface inspection is becoming a yield improvement tool.’‘

Designing machine vision systems that can inspect products moving at 3000m/minute is no easy task, but it is one of the benchmark achievements that have led to greater acceptance of automated inspection and process control systems in high-speed web inspection applications. ‘‘One specialty film customer not only realized a labor savings by going to automated inspection, but by taking manual labor out of the loop, they were able to increase the speed of their line by 85%, and they expect to reach a 100% speed improvement soon,’‘ says Vic Wintriss, president and founder of Wintriss Engineering Corp. (San Diego, California), a designer of web inspection hardware, software and systems.

Hard-won success in web inspection
Experts also attribute greater acceptance of vision technology in web inspection to falling vision system prices, machine vision system suppliers' and integrators' growing knowledge of web processes and manufacturing, and the ability to collect, analyze and present data in ways that operators and production machinery can use to optimize processes.

‘‘We can inspect at faster speeds in paper, for example, and still get good images,’‘ says ISRA's Goeckel. ‘‘If you want to look at defects, you can have the operator look at it and see the defect in a clean image. When you allow the customer to report and store all data from all production, you can query all sorts of data related to comparisons with previous orders, previous rolls, and previous month's production – compare any operational parameter to any other, all of which helps you understand the process and potential defects. Advanced defect classification has become important to many companies to better manage their processes and head off problems.’‘

Lastly, in today's litigious environment, defect records can help avoid unnecessary litigation. ‘‘The only way to guarantee quality is to do 100% inspection, and you can only do that with a vision system,’‘ says ISRA's Goeckel. ‘‘And in today's society, it's good to know what you're sending out. A $10,000 material issue can result in a $5 million claim for the end-user. If you know exactly what you sent out, you can show that your material wasn't at fault. Machine vision can deliver some pretty powerful evidence in support of your position.’‘

Quality vs. Comprehensive inspection
Vision system designers typically use two different types of vision platforms to inspect non-pattern and pattern webs: namely, line-scan inspection systems for 100% inspection of non-pattern and pattern webs, and area-scan inspection systems for high-resolution sampling of pattern webs for process control.

ISRA, i2S and Wintriss Engineering are examples of integrators that use line-scan cameras with sizeable memory and processing platforms to handle the high-bandwidth data streams for non-pattern web inspections. ‘‘We've had a lot of success using our WebRanger smart camera in metals, films, non-wovens, and paper, with paper being the biggest market,’‘ explains Pete Burggren, director of the WebRanger product line at Wintriss. ‘‘Anti-glare and anti-reflective thin films on LCD and plasma TV panels are another growing area.’‘

Line-scan vision system hardware and software designer i2S has developed a special line-scan camera called the CIS, affectionately nicknamed the ‘‘camera on a stick,’‘ that can be constructed in lengths of up to 2m with 200 dpi. Designed as a self-contained line sensor, the camera can be used for either pattern or non-pattern inspection when combined with i2S' PrintScan or FlawScan software, respectively.

‘‘Retrofits of existing manufacturing equipment are the largest application area for web inspection systems compared to OEM says,’‘ explained i2S's Borges. ‘‘A big problem with retrofits is space because the inspection system is usually the last thing put on the line. Many times, you have to re-engineer the line to accommodate the vision system front end, but the CIS sensor's size, flexibility and narrow profile can make integration much easier.’‘

Advanced Vision Technology Ltd. (AVT, Hod Hasharon, Israel), a specialist in the inspection and automated process control of printed or pattern webs, al so uses line-scan cameras for 100% pattern web on the back end of the printing press. ‘‘We use area cameras for process control and press control and line scan for 100% quality assurance,’‘ says AVT's vice president of marketing, Gal Shamri. ‘‘In the printing industry you see both.’‘ We even have a combined solution that incorporates the two technologies together offering the best of the two worlds.’‘

Unlike 100% web inspection, process control is ‘‘all about finding defects and quality variations in the early stages,’‘ according to Shamri. Using high-sensitivity area scan cameras with fast shutter speeds, combined with algorithms that ignore normal web variations and movements while highlighting defects, the high-resolution area scan cameras provide images that maintain high-resolution at high-magnification. This allows the system to perform print defect detection, verify bar codes or perform in line color measurement, at high speeds, while feeding valuable real time information to the operator for pro-active process control.

Using area scan cameras for registration and pressure control, Shamri estimates that printing press operators can reduce initial set up of a web from 30 minutes and 2500m of wasted material to a few minutes and a few hundred meters of material. ‘‘We use special targets to analyze the register and pressure of the different cylinders and adjust it automatically,’‘ Shamri says. This task can be done only with area camera technology.

The success of machine vision is forcing many OEMs of thin-film and web manufacturing equipment to include vision in their newest product offerings. In the meantime, however, helping manufacturers find ways to deliver improved products using existing equipment represents a large market for machine vision.

‘‘In the general market atmosphere, margins are tough, cost of material is getting higher and customers want their quality regardless of how much it costs the vendor, so quality itself becomes the issue. People are looking for technology that will help them deal with the cost. It's not a regional trend, but a global one. Europe and the U.S. are leading, but we're seeing the same trends in Latin America, Asia and China. The printing industry is a global industry and people buy products from different countries, with infrastructure all over the world. Companies want to have the flexibility to move production from one site to another, while delivering consistent quality. Machine vision helps these companies do that.’‘