EVERY MACHINE builder would like to know what its machine control, instrumentation and electrical systems will be like in five or 10 years. The more we can plan for the inevitable changes that automation technology will bring, the better we can benefit from it when it evolves into practical machine control functionality.
One way to envision the future is to look at specific trends in automation hardware and software, but it might be better to first look at more general trends that will inevitably drive the evolution of machine automation.
THE MACHINE OF THE FUTURE WILL HAVE:
We know better than that. Machine builders constantly balance the need for new automation to improve machine performance, safety and reliability with those support and customer reaction issues. There are many machines operating on “older” technology, and they’re not doing badly. They’ll move to newer technology on their terms, not a vendor’s marketing urgings.
This article aims to highlight emerging technologies in a way that will confirm to many machine builders that they have a good handle on emerging trends. For others, it provides a needed “heads up” that tells them it’s time to start thinking about the effects of these trends on future designs.
Easy as One, Two, Three
Automation, electromechanical components and labor are the three elements needed to operate a machine. If we look at the relative advances in these areas, we can discern trends that will shape the future of machine automation.
One general trend--apparent for decades and always gaining strength--is the rapid advance of automation, instrumentation, and electrical components and systems. These components and systems improve constantly and rapidly in their price/performance ratio, size and reliability. The microwave oven-sized PLCs of not-too-long ago quickly gave way to today’s less-costly PLC with similar capability that fits in the palm of your hand.
These advancements are especially striking when compared to the lack of advancement in electromechanical components and labor.
Electromechanical components sometimes do advance, as with improvements in motors and linear motion systems, and in the reliability and precision of linear guides and ballscrews. But these components often regress in price vs. performance as exemplified by recent huge increases in the price of raw materials such as steel.
When electromechanical components improve, these changes usually are slow and gradual, and cannot match the quantum leaps that we often see in automation systems performance.
What about labor--that final key component needed to build and operate machines? Labor costs certainly are increasing worldwide. Even China now reports a shortage of skilled labor and skyrocketing wages.
Even when wages are stagnant, as in much of the developed world, labor cost continues its inexorable march upward from increases in non-wage costs such as health care, worker’s comp and pensions.
It seems certain that both labor and electromechanical components will continue to increase in cost, especially as compared to automation systems. In most cases, an increased use of one of these elements will decrease the need for one or both of the others. It follows that the one factor decreasing in cost relative to the other two will see more use. In one way or another, the future of machine automation will be determined by the relentless, basic logic of this equation.
A Basic Truth: Automation Replaces Labor
For many years manufacturing output in the U.S. has been increasing at about 5% per year. Demand for manufactured goods has been growing 3% each year. If inventory levels remain basically stable, then this argues that manufacturing employment has been decreasing by 2% per year. Manufacturing productivity growth perhaps is the key factor in recent U.S. economic success, especially as compared to other first world nations.
|FIGURE 1: LABOR SAVER|
This necking machine adds top and bottom trim to beverage cans. Belvac Production Machinery discovered substantial labor cost savings by using remote diagnostics to eliminate much travel to customer sites. Source: Belvac
One of the main reasons to automate is to remove the variability of operator actions. This can be particularly true of complex labeling operations (See Figure 2). “Automation reduces required operator interaction with machines,” says Greg Borsos, automation manager at Weiler Labeling Systems, Moorestown, N.J. “I’ve heard many stories about how operator errors caused recalls and major rework. Because of this, many companies are removing as much responsibility from the operator as possible and automating more.”
|FIGURE 2: SIMPLIFY THE COMPLEX|
This machine prints serialized numbers on labels and then performs character verification inspection prior to applying the label. This process is used for tracking product during lab trials. One of the main reasons Weiler Labeling must automate is to remove the variability of operator actions. Source: Weiler Labeling
Application examples include checking that a cap is on a bottle, that the date/lot code is on a label, that all of the weld nuts are on an automotive frame, or that the correct number of tablets are in a blister pack. Further, part-feeding applications often require complex mechanical and sensor setup to ensure parts are fed in the correct orientation.
Close the Loop
Another way that automation replaces labor is through closed-loop control. Open-loop control requires constant manual observation and adjustment. Closed-loop control not only reduces labor, it also improves operations.
AutoJet Technologies, Wheaton, Ill., is the turnkey systems division of Spraying Systems, a manufacturer of high-quality spray nozzles. Autojet’s automated systems optimize spray performance to improve product quality and reduce production costs.
AutoJet makes extensive use of closed-loop control. “Manufacturers use closed-loop control to create smart production machines that can adapt to changing conditions,” reports William Kohley, vice president at AutoJet. The company provides closed-loop control with its gas conditioning system, which is used to cool hot gases and reduce gas volume at pulp and paper mills, steel mills, power plants, waste incineration facilities and cement plants.
The operator specifies the desired gas temperature; the spray controller, using temperature sensors, constantly adjusts variable frequency drive pumps to control liquid flow. “This closed-loop regulation mode is capable of generating consistent drop sizes and holding very tight temperature tolerances for peak system performance with minimal operating costs,” states Kohley.
Machine builders enjoy another big labor saver for their automation systems when they use standard hardware and software. The PC WinTel standard now is so well-known that there is little or no learning curve for machine builders and their customers.
Fadal Machining Centers, Chatsworth, Calif., manufactures vertical and horizontal machining centers for applications ranging from medical to aerospace industries. Manager of electrical engineering Khosrow Ansari envisions a PC-based control system with a standard bus communication. The PC would use PLC-type I/O configuration and measurement devices.
He feels that would allow Fadal to standardize on control system hardware and develop proprietary software. “This would open up a host of options that could reside or be connected to the control system such as CAD/CAM, coordinate measurement machines, and remote diagnostics,” says Ansari.
Coordinate measurement machines allow inspection in operation without removing the part from the machine. According to Ansari, this is not a very common application, but it’s growing. He says quite a few end users use probing, but a measurement system often would do a better job.
Another machine builder that sees the light regarding off-the-shelf Windows-based hardware is Orthodyne Electronics, Irvine, Calif. Orthodyne is a world leader in the manufacture of manual and automatic ultrasonic wire bonders for the hybrid and semiconductor industries (See Figure 3).
|FIGURE 3: STANDARD PARTS MAKE BETTER PARTS|
This large wire wedge bonder benefits from standardized automation components since the system often is integrated into automated production lines and needs a highly flexible networking architecture. The machine also includes an automotive vision system and operates on a two-level operating system. Source: Orthdyne
Automation Replaces Electromechanical Components
If electromechanical components are becoming more expensive relative to automation systems, the machine of the future will use automation to replace these components.
Machine-mountable controllers, I/O and sensors are a good example of this trend. Machine mounting eliminates the need for enclosures and associated hardware.
Another prevalent trend is replacement of electromechanical drive systems with servo motors and drives. These alternatives can eliminate, or at the least enhance, a host of components such as belts, gears, and other mechanical linkages.
Fiberoptic sensors replace entire classes of electromechanical instrumentation. These sensors are smaller, cheaper, more durable, and can operate at much higher temperatures than electrical and electromechanical sensors.
Unlike electrical sensors, fiberoptic sensors are not susceptible to electromagnetic interference and can therefore be reliably used in industrial machinery where electrical noise abounds (see "Getting Enough Fiber? sidebar below).
Another automation component that displaces electromechanical devices is the graphics terminal. Gauges, thumbwheels, and pushbuttons typically are eliminated when a graphics terminal is used, and this changeover has been proliferating as these terminals decrease in size and become cheaper.
The Machine of the Future
The direction seems clear. Machines of the future will use automation to replace labor and electromechanical components—a process that started many centuries ago. Examples of this abound, and more will crop up as automation becomes cheaper in relation to labor and electromechanical components.
Some of the specific automation technologies enabling labor and electromechanical component replacement are clear and present. Others technologies such as fiberoptic sensors are in the R&D phase. Still others have yet to be discovered.
As automation takes a bigger piece of the machine pie, the control professional’s job will increase in relative importance. Machine costs used to be dominated by electromechanical components, and these machines were very labor intensive to build and operate. Most of the cost of future machines will consist of automation components that promise to be ever cheaper, more durable, and higher performing.
Automation Drives Retrofit Trend
ELECTROMECHANICAL components might not be improving rapidly in terms of price vs. performance ratios, but they certainly last a long time. This creates an opportunity to improve machine performance via upgrades to automation systems. It also means that some of the machines of the future might actually be today’s machine with upgraded automation.
System integrator Concept Systems, Albany, Ore., often retrofits machines with new automation systems, and its customers see many benefits to this approach. “Often an old machine can be upgraded to produce more by cutting changeover times or decreasing maintenance downtimes simply by replacing an old, inflexible, hard-to-maintain-and-upgrade control system with a new one,” says Ed Diehl, principal at Concept.
Concept Systems replaced the Bailey DCS on this steam turbine and generator with a redundant Allen-Bradley ControlLogix system and a Rockwell Automation RSView HMI. Source: Concept Systems
“The old control system was hard to troubleshoot and its wiring was in bad shape,” says Diehl. “It was very expensive to call the original control system vendor in for service, and it was hard for Collins' own maintenance personnel to service the system.”
Concept replaced the old control system with a new Allen-Bradley ControlLogix system with redundant power supplies and processors. New graphical operator displays were added and many more alarm points were connected so that system status was easier to monitor and problems were able to be more quickly identified and diagnosed. System downtime was greatly decreased and responsiveness to boiler controls was improved, allowing the plant to remain on-line more consistently, making the power generation operation much more efficient.
Automation vendor Delta Computer Systems also delivers value by replacing automation systems on existing machines. “We often retrofit new closed-loop electronic controllers to old machines and material handling systems,” reports Steve Nylund, Delta's CEO.
“These older machines work well mechanically, but are either productivity-limited by manual or electromechanical controls, or have become maintenance problems due to outdated controllers or normal wear-and-tear on system components. New closed-loop controllers, instrumented with the latest position and pressure transducers, can make older machines function as well as or even better than new machines at a fraction of the cost,” adds Nylund.
Outsource Labor to the Machine Builder
THE LABOR expended in a factory to commission and start up a machine is extensive and expensive.
If the labor is performed by the machine builder’s own personnel, there are travel and lodging expenses to pay beyond the labor costs. If the labor is performed by the manufacturer’s automation professionals, extra expense is incurred because the manufacturer is not as familiar with the machine as the machine builder.
Because of these factors, Battenfeld Gloucester uses automation to allow customers to “outsource” labor to it.
Battenfeld Gloucester Engineering, Gloucester, Mass., designs, builds, installs and services customized plastic extrusion systems and controls.
“We use distributed I/O and sensors to create more intelligent control systems,” says Paul Brancaleone, engineering manager of software/controls at Battenfeld. “Our machines are becoming more modular, and this increases our internal assembly and test times.”
Because the extra time involved in producing modular machinery results in savings on the installation and commissioning side, Battenfeld customers see equipment go into their plants and be up and running much quicker, adds Brancaleone.
Battenfeld customers also outsource automation labor through remote access and diagnostics. “We can connect to the control systems remotely to help in diagnosing problems,” says Brancaleone. “This means that our customer’s equipment operators require less skill and time as the intelligent control systems remotely warn us when something is going wrong.”
Another way to outsource labor from the manufacturer to the machine builder is through quick-connection systems. Nearly all connector companies provide many types of quick style connections—ranging from spring-clamp to insulation displacement terminals to metric cords and connector sets--as opposed to the traditional screw-down components. “Decreasing the amount of time it takes to assemble a piece of equipment allows machines to be re-assembled quickly at customer sites,” says Todd Desso, product manager of industrial connectors at Phoenix Contact.
Getting Enough Fiber?
ACCORDING TO a recent article in The Economist, the most common type of fiberoptic sensor in use today is called a “Bragg grating,” which is the fiberoptic equivalent of a strain gauge. A Bragg grating is a region of a fiber where the refractive index has been modified so that it varies in a precise and periodic way, causing the grating to reflect light of a specific wavelength.
As the fiber is stretched or compressed, the reflected wavelength changes accordingly, and strain can be determined. Changes in temperature also change the fiber's properties in predictable ways. By incorporating several Bragg gratings into a single fiber--each tuned to reflect a different wavelength--it is possible to measure the variations in strain or temperature along the fiber's length.
Professor Anbo Wang and his employer, Virginia Tech University in Blacksburg, Va., created the Center for Photonics Technology to investigate and create sensors that can be used in harsh industrial applications.
The center has produced the world’s smallest known pressure sensor. The sensor is a mere 125 µm in diameter and can function at temperatures as high as 700° C. Competing sensors are limited to 500° C.
Developed by Yizheng Zhu, a Ph.D. student at the center, the sensor is fabricated directly on the tip of a fiber by micromachining and thermal fusion, giving it the same thickness as the optical fiber.
The sensor has minimal cross-sensitivity to temperature, resulting in a simplified sensor system with a wide temperature range. Its small size and low mass give the sensor an extremely high resonant frequency, resulting in a flat response over a very wide range of frequencies.
Sensitivity can be adjusted for different applications, with a pressure range as low as a few psi or as many as 10,000 psi. The sensor is another step toward the Center’s goal of developing pressure sensors that can operate above 1,000° C.