Subscribe and listen to AW’s podcast!
Subscribe and listen to the Automation World Gets Your Questions Answered podcast!
Listen Here

Optimizing the Factory Floor

Advances in software help engineers and managers to boost the efficiency of their operations.

Making quality shingles economically requires ElkCorp to collect 6,500 pieces of information every few seconds. Only modern manu
Making quality shingles economically requires ElkCorp to collect 6,500 pieces of information every few seconds. Only modern manu

While getting ready for an important meeting, systems engineer Paul Rogers got a call from the line producing the laminated asphalt shingles that Dallas-based ElkCorp makes. The continuous web of shingle material had broken somewhere in the finished product looper, a device that snakes the web around a series of rollers to create a buffer between the cooling and cutting stations. Because downtime costs the company thousands of dollars an hour, Rogers would have to drop everything and even miss his meeting to find the cause, while the line workers rethreaded the web.

As an increasingly larger number of his counterparts at other manufacturing facilities can do now, thanks to advances in software, Rogers logged immediately onto the plant’s computer network to gain access to the company’s manufacturing execution system. Using IndustrialSQL Server from Wonderware, of Lake Forrest, Calif., he looked at the historical trends for tension. He noticed that the electrical-current load on one of the motors had been climbing steadily until the web broke. He relayed the information to the mechanics, who discovered and replaced a worn bearing.

“Within five minutes, we were able to look at our entire line, identify the problem, and go right to it to fix it,” says Rogers. And, he made his meeting, which would have been impossible in the past. Diagnosing the problem could have taken hours to days because the line is quite long, approximately a football field and a half. While engineers and mechanics pored over the line, the operators would relace the looper and run the line until the web broke again.

The key for unlocking this and other gains in operating efficiency has been advanced data-handling technology. “The trend over the last four years has been the convergence of older manufacturing technology and new information technology (IT) that has been used on the administrative-side computer systems,” observes Rogers. “Management brought me [from the IT department.] to the engineering department so engineering could take advantage of technology used on the administrative side of networks.”

Not only has the ubiquitous transmission control protocol/Internet protocol, or TCP/IP, migrated to the plant floor, but other de facto standard tools also have worked their way into manufacturing software. Microsoft’s structured query language (SQL), for example, is the key technology around which Wonderware organized its real-time database. ElkCorp uses IndustrialSQL Server to collect about 6,500 pieces of information from the line every few seconds and store them in the database for two years.

Maximizing uptime

Software in the Wonderware suite then can use the data either immediately or later. Some, such as InTouch, retrieve and analyze pertinent information in real time to display current operating information or to trigger alarms. Other software helps engineers to analyze historical data not only for troubleshooting broken webs and other problems, but also for optimizing the process. “When we have a particularly good or poor run, we can check the tensions, the temperature, percents of the raw materials, and other parameters to recreate the success or to avoid the problem,” says Rogers.

The massive database also helps engineers to optimize production in another way. They can prepare for regular maintenance shutdowns by analyzing operating data to identify any developing problems for correction while the line is stopped. During one of these analyses, Rogers noticed that one of the catchers collecting cut product for wrapping had jammed 25 times in two weeks. Because the other catchers had jammed only once or twice during the same period, he added the catcher to the list for maintenance during the next shutdown.

These kinds of analyses also can uncover ways that redeploying resources can improve productivity. Consider the experience that an automotive supplier of clutches had with software from Rockwell Software, a unit of Milwaukee-based Rockwell Automation Inc. Assembly machine and systems builder Weldun International Ltd., of Bridgman, Mich., had integrated the software into a production line containing approximately 40 automatic stations and ten manual stations for making the clutches. Management at the supplier could analyze the contributions of each operator to the line, once the equipment had been running for a while.

An analysis revealed that some operators performed certain tasks better than others. “So by simply relocating people around the line, management increased production dramatically in a matter of a week,” says Mike Zimanski, an applications engineer at Weldun. Discovering this simple way of increasing efficiency was nearly impossible before, because no way existed for compiling the information collected manually.

Hey, this pays

Despite the dividends that manufacturing execution software pays in uptime and efficiency, many manufacturers felt in the past that the productivity gains had been too small to justify the expense of buying, installing and maintaining the software and the infrastructure to run it. The situation has changed over the last four years, however. A growing number of these manufacturers is reconsidering the economics, either because they need other benefits that the software offers or because lower costs of implementation have changed the cost-benefit ratio.

Traceability is one benefit that has become increasingly more important. “The past three years have seen a huge demand for records of the process and the product as it’s being built,” reports Zimanski. “We haven’t built a machine in the past two years that hasn’t had to track a complete history for every part built.” Weldun systems perform this task with the RSBizWare suite and the FactoryTalk architecture from Rockwell Software, which allow the entire control system to share the required data.

Demand for traceability is particularly acute in the medical and automotive industries. Governmental regulations require traceability in the medical industry, and the Big Three automakers are pushing the concept, forcing their Tier One suppliers to assume the responsibility for the costs of warranty and recall. “So the software and process information have become a kind of insurance policy,” notes Zimanski.

Communications standards also help to cut costs. Most industrial equipment can talk to Ethernet now, and standards such as OPC, an open communications standard, and ISA 88, a batch processing standard promulgated by the Instrumentation, Systems and Automation Society, specify a universal way for establishing the free flow of information among disparate platforms in real time. Custom drivers are no longer necessary. To streamline the flow of information on its production line, ElkCorp makes use of these standards. OPC is especially important for the company’s manufacturing operations because the line contains a half dozen small networks, or control loops, each of which controls a portion of the equipment along the line producing its shingles.

A recent upgrade of the communications infrastructure at ElkCorp is only making the need for communications standards more urgent. “We’re replacing some of our older drives with new Siemens drives and Profibus fiber-optics network,” explains Rogers.

To avoid any hitches in the flow of information among such disparate networks, manufacturers expanding or improving their equipment or facilities can follow ElkCorp’s lead by investing in manufacturing execution software that has up-to-date communications tools. “You want to have your eyes on the entire line,” says Rogers. “So you need to make any connections between networks seamless so engineers and managers can see what’s happening” and the equipment can exchange information.

This ability of modern software to reach across a network and extract information from machinery and to document product quality even extends to the computer numerical controllers (CNCs) found on machine tools. The closed nature of these specialized controllers has made mining them for information historically more difficult than extracting data from programmable logic controllers (PLCs). Most manufacturers of CNCs, however, have gotten around that problem by integrating personal computers (PCs) into their products. This tactic gives them a standard front end that can link to the factory’s computer network seamlessly, yet preserves the proprietary architectures and dedicated processing that machine tools need.

Because quality is a must in jet engine production, GE Transportation’s Aircraft Engines Group was quick to exploit the communications tactic. The organization uses it to verify and document the quality of the jet-engine components that it grinds on three Campbell CNC radial grinders in its Raleigh-Durham plant in North Carolina. Monitoring and documenting the process involves gathering more than 600 parameters during the grinding operation to show that tolerances are within ±0.001 inch, and to comply with the company’s Six Sigma manufacturing philosophy.

Grinding is slow, taking about an hour per feature. So depending on the type of engine, the grinding of one assembly—such as an array of turbine components for an engine’s hot section—can take from 12 to 24 hours. “Jet-engine assemblies are made of high-temperature, heat-resistant metals that are difficult to machine,” explains David Smither, an assembly engineer at the Aircraft Engines Group’s Cincinnati office.

After each machine grinds a feature, it retrieves a touch probe from the tool changer and measures the feature automatically to verify that it matches the blueprint. The data go to a PC running the Cimplicity HMI (human-machine interface) from GE Fanuc Automation Inc., an affiliate of GE Infrastructure, based in Charlottesville, Va. The data flow over a network that connects the PCs and GE Fanuc Series 15B CNCs with fiber-optic high-speed serial buses and about 150 feet of fiber-optic cable. “With this system, a direct, high-speed, fiber-optic link connects the CNC with the PC,” says Smither. “Data transmission is instantaneous.” Past technologies required a 10-minute wait per feature.

Not only does the HMI provide access both locally and globally to operating data within the CNC and a means for editing and managing part programs, but it also lets operators search the database, compare ground features against the blueprint, and print reports to track quality control. Smither even has access to the data from his office hundreds of miles away in Cincinnati.

“We have a window into the process,” notes Smither. “It’s a whole new level of data access and programming convenience for CNCs that improves quality and productivity.” In other words, the advanced software helps him to optimize the factory floor.

For more information, search keywords “MES” and “software” at www.automationworld. com.

Discover New Content
Access Automation World's free educational content library!
Unlock Learning Here
Discover New Content
Test Your Machine Learning Smarts
Take Automation World's machine learning quiz to prove your knowledge!
Take Quiz
Test Your Machine Learning Smarts