The last few years have not been exactly the best of times for manufacturers. In fact, they have been the worst of times for many, especially those not invested in flexible automation and information technology. As the recession proved rather painfully, concentrating on controlling the cost of labor is not a winning strategy in the developed world. This strategy gives the third-world countries the upper hand.
The only way for companies in developed societies to prosper in the long run is through a combination of ingenuity and investment in capital equipment that allows them to become more productive, yet retain the flexibility to roll with a changing market. In other words, success means deploying flexible automation and information technology. It means getting a good return on one’s investment in automation.
Recipe management
One company reporting a healthy return on automation (ROA) during these times is Coty US Inc., a privately held manufacturer of bath and body creams such as Calgon and the Healing Garden. Based in New York City, the company invested in an automated information system at its 906,000-square-foot plant in Sanford, N.C., to manage nearly 500 recipes and control the quality for producing more than 300 products that it had outsourced to a variety of subcontractors in three states. It found that outsourcing is not necessarily the best business decision.
By consolidating production into one facility fitted with the latest information technology, the company was able to increase production by 500 percent, save $4.1 million a year in operating costs, and improve product quality and consistency. With just two operators per shift, the plant can process up to 60,000 kilograms of product in 24 hours; whereas, the contract manufacturers could produce only 15,000 kilograms during the same period. Moreover, “we not only get the consistency we want, but we’re also able to certify it through electronic documentation,” says Joe Kennedy, director, application development.
The source of this success was the three-stage production process designed by Geometric Controls Inc. (GCI), and the way in which the systems integrator from Bethlehem, Pa., linked the stages electronically to track material usage. Barcodes and custom software running on control, communications and visualization technologies from Rockwell Automation Inc., Milwaukee, allow management and buyers to view what is in stock, where it is and whether it has been tested. The system not only ensures that enough material is on hand to mix each batch properly but also verifies that operators mix the ingredients in the correct proportions.
As raw materials arrive at the plant, an in-house laboratory tests samples and reports the results to the plant’s Oracle enterprise manufacturing system. Accepted materials remain in storage until the enterprise system generates a work order and bill of materials. When it does, it sends them via an Ethernet network to the pre-weighing station, where software that GCI developed scales the recipes and bills of materials for the work order and generates the necessary instructions. The operator then weighs the ingredients, puts them into containers, affixes a barcode label generated by the software and sends the containers together to the processing mezzanine on the same pallet.
At the processing mezzanine, the operator decides which of the six kettles to use for the jobs, assessing their availability using an RSBizWare batch-management server from Rockwell Automation. Once the operator selects a kettle, the batch-management software manages the heating, cooling, mixing and blending inside the kettle for each phase of the process detailed in the recipe. While adding the ingredients, the operator scans the barcodes, enabling the batch management software to monitor the ingredients and verify the consistency of each batch.
When a batch is ready, a Rockwell Allen-Bradley ControlLogix controller directs it automatically to one of three high-speed tube-filling lines or to one of three high-speed bottle-filling lines connected to the kettle. In the old days before the consolidation, operators at the subcontractors had to connect the kettles to the correct line manually. Consequently, batches often sat in the kettles until the operator found the time to do it and begin the packaging operation. By automating this process, Coty not only eliminated this and other inefficiencies, but also has poised itself to capture a healthy share of a market expected to grow to $1 trillion by 2010.
Watch for trouble
Automating production is not the only way to generate healthy returns. Monitoring equipment for predictive maintenance can pay handsome dividends too. Consider the experience of Millar Western Forest Products Ltd., in Whitecourt, Ab., Canada, one of the world’s largest suppliers of bleached chemi-thermo-mechanical pulp. Since installing a high-speed data acquisition system from Binnington Development Corp., in Vancouver, the company is reaping $15,000 more in product every day from its Sunds Defibrator hydraulic slab pulp press.
Initially, management undertook the project to simplify and improve maintenance of a notorious bottleneck. Because the slab press is 90-feet high by 20-feet wide to compress heavy, snowflake-like pulp into 0.5-cubic yard bales that weigh about 500 pounds, its size and complexity made diagnosing problems difficult. The only means of collecting diagnostic information was through diagnostic ladder logic that only its programmer could decipher. Consequently, information was often either unavailable or impractical to obtain, and operators and maintenance technicians often were reacting to symptoms rather than the root causes of problems.
“We had ideas about how we could improve the performance of the press,” recalls Rod Savoy, an electrician at Millar Western. However, for fear of putting “too much stress on the hydraulics, we didn’t dare test our theories until we had more details about the machine’s performance.”
Since Binnington Development fit the press with a custom data-acquisition system, maintenance technicians at Millar Western now can see tiny flaws in the process that had been impossible, or at least impractical, to find. Even though some of the flaws are slight and might seem insignificant, the company has learned that finding and correcting them before process problems occur has not only lowered downtime but also increased throughput by 10 percent.
“Before we started using this high-speed data acquisition system, the press took 45 to 48 seconds to create a bale,” explains Savoy. “Now, it takes only 38 to 40 seconds.” Because the plant runs around the clock, the boost in productivity adds up to an unexpected dividend of $15,000 per day.
The ability to collect the information necessary for maintaining the press and optimizing its performance comes from the high-speed data-acquisition system that Binnington Development built from Cimplicity human-machine interface software and Series 90-70 programmable logic controllers (PLCs) from GE Fanuc Automation Inc., Charlottesville, Va. The data are logged onto an SQL-server database at 60 samples per second, which is a 60-fold increase over typical polling and collection rates. The result is a system that can peer deeply into the process without interfering with it.
Although calculating an ROA is common practice in large companies, not everyone in small and medium-sized companies thinks that the exercise is worth the trouble. Steve Weyreter is one of them. Like many entrepreneurs, he prefers taking the time to make informed decisions and then moving on. Consequently, he doesn’t know what the return is with any real accuracy on the $8 million investment that he made in an automated machining cell at Major Tool and Machine Inc., a contract manufacturing company in Indianapolis.
Built for high-mix, medium-volume production, the cell is the third largest that machine tool builder and systems integrator Cincinnati Lamb, Cincinnati, has ever constructed. With room for three more machines, it currently has nine workstations: four five-axis machining centers, one four-axis machining center, two vertical turning lathes (VTLs), a deburring station, and a coordinate-measuring machine (CMM). The machines sit on both sides of a 180-foot rail, on which a flat-topped vehicle travels to shuttle piece parts among the nine stations. Castings, sawed blanks and semi-machined workpieces are machined, deburred and measured without human intervention once they enter the cell.
Part of Weyreter’s reluctance to calculate a return on automation comes from the fact that he doesn’t know where to start counting. He had been buying the machine tools as stand-alone equipment over the years, acquiring them as the company needed more capacity to produce the metal piece parts that it makes for the national laboratories and the power generation, aerospace and defense industries. He organized the machines into the automated cell only when he had to uproot them to move them into a new building. At that time, he bought a couple more machines and the rail-guided vehicle (RGV) from Cincinnati.
Complicating the calculations further was that the integration of the equipment occurred in two stages. Cincinnati Lamb installed the RGV and connected to it the five machines that it had sold to Major Tool. “Then we took the cell one step further by integrating our VTLs and our CMM,” says Weyreter. “Having a cell that just did milling wouldn’t get the job done.” His engineers also linked the cell to the enterprise resource planning (ERP) system.
“When you do it this way, how do you account for the costs?” he asks. “Do you go back and look at what we paid new for the machines four or five years ago? Or do you just look at what we put into the cell controller and RGV?”
Another difficulty in calculating an ROA is that the business is contract manufacturing, both machining and assembly. “As a job shop, we don’t have a product,” he explains. “We have medium-volume, repetitive jobs, but every month we generally machine different parts.” So comparisons are difficult because the cell has few opportunities to cut parts that were done on the machines when they were stand-alone units.
Despite the difficulties in calculating an ROA, Weyreter knows roughly what it is. Being a successful contractor, he knows how much more productive the cell is than stand-alone machines. “The cell has given us more spindle time,” he says. “By reducing the overall cycle time on each machine, it’s given us the ability to run 20 percent to 25 percent more parts” in a given amount of time. Moreover, the mix of machines and the flexibility of the automation allow the cell to cut any mix of 32 parts at the same time—whether they are the same or different and whether they are prismatic or cylindrical.
Directing traffic is a Cincron cell controller from Cincinnati Lamb. Using the routes for each of the 32 parts flowing through the cell, the cell controller then determines the optimal route for the RGV to deliver all of the workpieces to the machines so the operations occur in the correct order and in the shortest time possible. At each station, the machine’s controller downloads the necessary part programs and orchestrates activity there.
After each machine tool completes its task, it checks the part with a probe, uploads the data to the cell controller, and sends the part to another machine for further processing or to the CMM for final inspection. Meanwhile, the cell controller collects all of the measurement data and route information and uploads them to the ERP system.
To Weyreter, the efficiencies in the flow of both parts and information that come from organizing the machines into an automated cell are priceless. So the ROI calculation doesn’t matter, just the returns that Weyreter’s business has gained. They not only are high but also have contributed mightily to the good times that his company has enjoyed through the recent bad times.
See sidebar to this article: Adapt Your Accounting Practices To Exploit Automation
Leaders relevant to this article: