By Phil Burgert
Better data acquisition technologies for factory settings are raising interest among machine builders and their customers. Maintenance and service top the list of benefits that can be leveraged, but increasing capabilities for enterprise resource planning (ERP) systems also are coming along.
Data collection capabilities now being added include remote diagnostics, wireless connections, performance payment management, real-time control, heavy-duty mathematics and other developments. Machine builders say their customers are asking for these capabilities more frequently.
Current technology simplifies and defines a better line between long-term support needs, says Mark Chrismer, senior project engineer at automated material handling systems company FKI Logistex, Leicestershire, England. It also provides opportunities to use common programming languages and reuse existing code.
Software advances in support of common languages like .Net expose connectivity in new ways that help reduce cost. says Chrismer. Systems are becoming more readily connected to corporate networks. This is a key difference as information technology and control technologies become more compatible through the new advances, while support costs are reduced as a result of the improved connectivity.
FKI Logistex, he notes, prefers to work with single source providers in developing and building data acquisition systems in which the provider is required to build features and capabilities into its products. Chrismer says his company has been attracted to products from Wonderware that make use of data collection, storage and retrieval technology advances allowing database professionals to perform long-term support of features that previously required controls engineers (Figure 1).
In some cases, data acquisition also helps a machine builders customers use pay-for-performance type arrangements when a machine isnt purchased outright, but is compensated based on uptime or the performance of the machine, he says.
In addition we find that the OEM wants to back up that requirement with some benefits or value to the end user, as well, says Bradshaw. They look to provide some sort of reporting capabilities to the end user rather than just typical HMI, graphical interface and squiggly line trend information.
Much of the data acquisition in manufacturing now is becoming more machine-specific, notes Bradshaw. People want to get more out of each machine, as opposed to an overall line. So the granularity of data becomes more apparent.
Machine builders, he says, strive to either increase sales opportunities or decrease costs because the market for brand-new machines is not as buoyant as it has been and they are looking for other ways to grow their business. Options include increasing add-on capabilities and reducing the cost to maintain the machines and/or provide warranty coverage.
If they can do remote diagnostics and solve a problem without having to send somebody to the site, theyve saved an absolute fortunethousands of dollars on one machine in one downtime, says Bradshaw.
Both Bradshaw and Chrismer downplay the role ERP systems playi in driving the development of data collection systems, although ERP might have an increasing role in machine data collection in the future.
I think the users of ERP systems drive the needs for increased reporting, says Chrismer, adding that hes noted no direct connection between the need to transport or import data for reporting purposes and the development of data collection. This information becomes part of the process of decision-making in ERP.
Bradshaw agrees. The greater opportunity actually is providing better uptime for the machine because, at the end of the day, the machine is only making money while it is manufacturing products. Bradshaw also notes that in the next few years ERP will increasingly push into machine supervisory systems. Wonderware, he says, very infrequently is asked for ERP connections when selling software through a machine builder. You could probably count on one hand when a customer said that they must link in to SAP at the machine level.
That probably will change, says Bradshaw, claiming that Wonderware is working with some large packaging machinery companies on manufacturing execution-system (MES) data collection capabilities for machines.
What we are doing at the moment is not at the individual machine level, he says. Were more at the line level where we try to put in an MES data collection layer over a machine line rather than an individual machine, because there are very few lines that actually come from just one vendor. Theyll buy different machines from different vendors to make a line.
When current machine builder development efforts increase, data collection will push down to the machine layer, he adds, noting that ERP started off as a financial tool and moved into resource planning and gradually is working down to machines.
More and more on the machine side of it, the capabilities are getting more closely linked, he says. So I think its only a matter of time before that becomes more of a requirement. But today if we tried to push that side of it, we would have to be giving it away because the machine builder sees no value in it.
The Future Is Now
Andreas Somogyi, managing director of global marketing, ProSoft Technology, says hes seeing ERP systems in the mix to transport or import data.
ERP systems play a big role, he says. At the end of the day thats where all the data will end up and where the scheduling and supply chain and material handling requirements will come to the plant floor. Somogyi says some of ProSofts partners supply solutions for connecting programmable automation controllers (PACs) directly to MES and ERP.
Data acquisition capabilities are part of the way that machine builders go to market and do business in a world that is increasingly driven by globalization, cost of ownership and competitive pressure, adds Somogyi. They look for connectivity in data systems, which gives them an opportunity wherever that machine goes globally, he says. There might be an OEM in Germany shipping to the U.S., to Europe and to Asia-Pacific. They need a level of standardization so they can connect to their machines globally for service capability, for maintenance and, through the development process, they can make sure the data acquisition devices on the machine provide such functionality. Somogyi claims ProSoft makes such standardization affordable by providing communication interfaces that allow data communication interfaces between controllers from multiple companies.
Companies such as ProSoft Technologies provide a gateway solution between those different technologies, and between different PLCs, he says. Thats a standalone gateway solution outside of the chassis. But we actually go a step further and do in-chassis solutions, which go with a communication module that plugs directly into the control logic or a logic platform product from Rockwell Automation, for example, or it goes directly into a Schneider Electric Quantum PLC.
Cant Separate Data From Control
Data collection and control data are interwoven at Triangle Package Machinery, Chicago, which uses an in-rack PC to handle high-speed calculations for control of all functions of a multi-head food product package weighing machine.
On a typical machine with 18 chambers, approximately 250,000 weight combinations are possible, explains Steve Bergholt, chief engineer electrical/electronic. Thats way too many to do in a PLC, he says. Bergholt notes that his ProSoft solution supplies the required flexibility of a programmable logic controller and the processing speed of an industrial computer to manage the data combinations.
Another heavy user of collected machine data is Stijn Schacht, head of research and development for Vapo Hydraulics, Dadizele, Belgium. He uses a PAC from National Instruments in its response to customer requests for control of custom hydraulic systems, including a precision position and control system developed for a prefab concrete slab manufacturer as part of a storage system holding 100 tons of concrete (Figure 2).
Brian MacCleery, senior product manager for machine design controls at National Instruments, says his company has found that quite a few customers are using CompactRIO for such signal processing capabilities. There are a lot of diverse requirements for signal types that arent being well-served, he says, noting that custom interfacing can be used in such applications.
Get What You Pay For
Vendors say that among the easiest ways to keep data collection systems economical, easy to support and not terribly complicated would be to use off-the-shelf software, rather than trying to develop custom-made databases.
It is obviously far easier and, over the long-term, more cost-effective to use commercial, off-the-shelf historian-type products, says Wonderwares Bradshaw, noting that his company offers special license products that have certain restrictions but allow machine builders to reduce commissioning time or reduce maintenance time or fault diagnosis. But they also can pass some value to the end user as well, he says. This has the ability to give an end-of-shift or end-of-day production-run type report, but it doesnt go as far as giving a batch-to-batch or shift-to-shift or day-to-day comparison.
Bradshaw notes theres obviously value if the end user later decides to take advantage of the batch-to-batch, shift-to-shift and day-to-day comparison reports. He can do that simply by upgrading the license because all of the old history is still retrievable, he says. Its just that the original license was restricted to get to the price point. That functionality is not there originally.
Chrismer notes that the benefits of data acquisition are among the most difficult things to put value on. At one end of the spectrum you have the cost of doing nothing. We could all agree that ignoring real data and not applying it to your operations has a cost. At the other end is the cost of collecting all of the real-time information that you can and analyzing all of it.
He suggests that the truth lies between those extremes. Our applications tend to focus on near-real-time datalets say an hours worthand building comparisons of this data to planned performance, says Chrismer. This would be one measurement method. What is the cost of not knowing the variation to planned performance for the past hour?
Real measurements for cost or benefit can only be done within a specific operational context over time, he adds. The efficiencies or inefficiencies measured over time provide another measure, he says. With some amount of real data combined with presentation of data to operations and analyst personnel, an area of focus with a good cost/benefit potential can usually be found.
Philip Burgert is a freelance writer and frequent contributor to Control Design, specializing in the technical subject matter.
Data acquisition can serve many purposes. From machine diagnostics to exporting to an ERP system, the use of data is varied. What are your customers using it for. Join the discussion at ControlDesign.com/daq.