Bad news, as the saying goes, flows downhill. In that, it is very much like current trends in the consumer packaged goods (CPG) industry.
If you consider the shelf-space at the retail end of the chain to be the top of the hill, then pressure for quicker, more flexible response to consumer demand runs right through the manufacturers to their suppliers and on to the original equipment manufacturers (OEMs) that supply manufacturing equipment—OEM’s such as Covington, Ky.-based R.A. Jones & Co. Inc., which supplies high speed packaging equipment for Fortune 100 companies such as Kraft and Anheuser Busch.
When Darren Elliot, the control engineering manager for R.A Jones, was challenged by his superiors to deliver that responsiveness by cutting machine production lead time by more than half, he realized that a totally new approach to engineering design would be required.
Elliot knew that his existing approach—which was reliant on centralized controllers running line-shaft driven machines—could not possibly change things enough to slash 26-week cycle times to his target of 12 weeks. To accomplish this, Elliot chose to adopt a modular architecture from Rockwell Automation Inc., of Milwaukee, that would not only shorten machine lead times for customers, but also pay benefits by turning his pouchers and cartoners into plug and play elements of the manufacturing line. Now, eight months into the new approach, Elliot is already claiming success with some product lines and says he’s getting close with the others.
Motion-centric
As equipment has become more motion-centric—moving away from its traditional reliance on mechanical cams and drive shafts with only one or two motors, and instead adopting multi-axis servo motor systems—a huge amount of pressure has been placed on the controllers. Each axis of motion must be controlled by the programmable logic controller (PLC), which only has so much processing power. However, motion-centric design also offers the underlying structure that makes a modular approach possible. To start with, why not take the most basic control functions and place them right at the axis of motion?
“In the past, motion control had to be done from the controller level by a single controller,” says Al Morin, a spokesperson for Hoffman Estates, Ill.-based Bosch Rexroth Corp., one of a number of vendors offering modular solutions. “As you add equipment, it increases the workload on the controller, slowing it down.
“Decentralized control takes the intelligence and puts it right at the drive level. Now you can add components, achieve scalability and no longer have to worry about the controller. The drives do the overhead work, while the controller maintains overall management of the machine, giving you one point of overall control.”
Distributed intelligence allows for advanced functionality such as electronic line shafting, cam tables, high-speed registration and absolute feedback functionality to be performed by the drive—instead of the main processor. Additionally, it allows for complete scalability in axis count. Because the processing power is integrated within the servo drives, no additional burden is put on the main controller. Adding additional axes is as simple as adding in additional servo drives.
According to Craig Resnick, a research director with Dedham, Mass-based analyst firm ARC Advisory Group Inc., the packaging industry is leading the charge for this kind of plug-and-play architecture. “OEMs are using this approach across the board. Any application that involves OEM machinery, that’s where we’ve seen huge uptake.”
As manufacturers are driven to produce a wider variety of more complex products, changeovers will become more frequent and this will require more flexible, more modular, less dedicated machines. In the future, even entire lines will need to be easily reconfigurable, he says.
Ripple effect
Although the packaging industry is leading the charge for modular architectures, the plug-and-play approach is making a splash throughout the manufacturing world.
Force Measurement Systems (FMS) of Fullerton, Calif., offers one example of the acceptance of the approach. The company, a supplier of thrust measurement and control systems for rockets and jet engines for, among others, NASA, recently adopted a modular input/output (I/O) system from Opto 22, of Temecula, Calif.
FMS’s thrust measurement and control systems include a built-in calibration system that records analog load cell signals and many other value gradations. As part of a control loop, these values are then converted and used for relay outputs to motor controls and other digital (i.e., on/off, open/closed) outputs.
The digital phase of this control loop is done with Opto 22 I/O. A host of analog I/O, including many load cell input modules, is used for the measurements. These analog readings are then converted to digital signals and Opto 22’s dry contact digital modules drive relays to initiate motor startup and shut down, and perform on/off switching.
“Modular control systems are more flexible in terms of their being able to handle a large number of analog, digital and serial I/O points simultaneously on the same rack,” says Opto 22 spokesperson David Crump. “Just pop in the appropriate I/O modules as you need them. This translates into the ability to control a greater number of devices concurrently.
“Modular control systems provide significant cost savings as engineers have the option of purchasing only the modules they need, and can deploy a new system or augment an existing one a single rack at a time.”
According to David Bauman, technical director for OMAC—the Open Modular Architecture Control users group—evolving industry standards have been a critical underlying enabler for this kind of machine design.
“In the past, a lot of companies would typically design systems that had vendor standards for control systems. To make it easier to integrate them together, you’d have to buy everything from one vendor and its partners.”
On the other hand, industry standards that are supported by many suppliers mean you are no longer limited to using one vendor, which conveys more competitive pricing power to the manufacturer and makes the machines easier to integrate into the manufacturing line, which saves time and money.
For example, says Bauman, in the past, whenever an engineer purchased a PLC for a job, the I/O by had to be purchased from the same vendor. Most PLCs today use industry standard device networks such as Profibus and DeviceNet. Hundreds of vendors have products that support these device networks. Because the I/O cost is typically a large part of the cost of the control system, competitive I/O cost is critical to cost effectiveness.
Industry standards also influence engineering costs. “OMAC’s PackML standards for packaging language and ISA’s S88 standard for batch systems are two examples of this,” says Bauman, in reference to the ISA-88 batch standard, commonly called S88, promulgated by the Instrumentation, Systems and Automation Society. “The S88 standard defines a common language for describing batching systems, and is now built into many vendors’ control systems, he notes. “Batch applications using S88 are seeing engineering reductions of 40 percent or more.”
Meanwhile, implementation of PackML— which defines a set of naming conventions —has significantly reduced the engineering required to integrate a number of packaging machines into an integrated packaging line.
“These standards and guidelines will play a vital role in the horizontal and vertical integration of packaging operations, and packaging operations with processing operations,” says Resnick. “This in turn will reduce project time and time-to-market, increase production synchronization and flexibility, reduce inventories and enable improved continuous quality verification.”
Resist the urge
The modular approach offers great benefits in terms of flexibility and customer responsiveness, but R.A. Jones’ Elliot cautions against the urge to get carried away and install an intelligent control on every axis of motion on every machine.
Mike Wagner, business development manager, Global OEM Team, at Rockwell Automation, concurs. “You can be modular and still not be efficient. It really comes down to understanding your process and understanding the boundaries. I’ve seen people go to the extremes in terms of flexibility and end up with a machine that nobody could buy. Some OEMs were putting servo motors on every flap tucker, and you ended up with machines with 20 axes of servo control. It’s too expensive. It’s too big. No one’s going to make use of all that technology.
“Now the pendulum has swung back to the middle, where it belongs. Now they’ve really pared it down and figured out where they need flexibility and where it’s not important.”
It’s not necessary to swat a fly with a howitzer, adds Elliot. “Use the right amount of technology. For example: a line shaft application might be a perfect fit for a given job. “If you’re running only one size [of carton] on a machine, you might not need the flexibility of a servo-driven machine that can adjust to 40 sizes.”
For more information, search
keywords “modular” and “servo” at www.automationworld.com.
See sidebar to this article: Software Goes Modular
