A potential cure he mentions is third-party analytical assistance, such as that provided by ScientificConservation (www.scientificconservation.com), Berkeley, Calif. “What they do is take data from customers and normalize those data, and put those into a structured database,” Brodie says. For example, with heating, ventilation and air-conditioning systems, the Berkeley company could collect data from a chiller, which Brodie says has “tons of data associated with it.” Those data will then be put into a virtual piece of equipment, when “they can run models for different conditions.”Recent advances in computers with multi-core processors allow high-performance throughput, adds Vineet Aggarwal, product manager with vendor National Instruments Corp’s (www.ni.com) data-acquisition (DAQ) group, Austin, Texas. Multi-threaded hardware drivers and parallel-programming-development environments make it possible to analyze large amounts of data immediately, in real time, with signal-processing algorithms, he notes. Or those data may be streamed to a disk for offline analysis, Aggarwal adds.This broader, higher-speed data collection and mining goes a step beyond traditional DAQ, Brodie suggests. Algorithms now can be written to mimic an end-user’s process and, with regular data input, “they can find things such as stuck sensors, for example,” he states.In this beyond-traditional DAQ, databases will play a bigger role, Brodie forecast. “So anything the DAQ can do to be more effective, to reduce that cost of modeling data—that’s going to be the future.” Becoming more effective translates to going from flat files, he explains, “to essentially where the model will be on the DAQ and you’ll just be importing data.”Wireless DAQBesides DAQ’s advance into data collection and mining, Brodie sees much more interest in wireless; specifically, wireless remote input/output (I/O). “DAQ can now be put on pieces of equipment and moved around a facility.” That means no wires, and the flexibility of moving equipment to a new area and not losing data, he notes.But a problem surfaces because of the frequency of radio bands. “As you go up in frequency, you lose [signal] penetration, but you can improve collection,” Brodie states. More important than the band itself, though, is the need to have technology “to find your way back” to the receiver or home base. That need led to mesh and grid network systems, Brodie comments. “Mesh has come into its own in the past one or two years.” Mesh allows end-users to overcome “the big problem,” routing, which slowed DAQ systems, he explains.Finding home, however, means end-users need to know how many radios they have, in what group and in what area, Brodie explains. “You’ve got to leave a certain amount of radios within a certain radius in a sector to have a mesh.” That may mean end-users will have to install permanent towers. But the payoff is worth it. “You’ll start also to track your density in a sector using a signal alarm, similar to a high/low alarm.” Wireless is a very specific solution, though, that may not suit everybody, notes Steve Byrom, a product manager with Yokogawa. “It’s inherently costlier than a traditional wired system.” So why choose wireless I/O? “Because end-users need data they cannot get any other way.” And when they get it, they can listen to what those data say, before potential problems become serious.C. Kenna Amos, [email protected], is an Automation World Contributing Editor. Yokogawa Corp. of Americawww.us.yokogawa.comScientificConservationwww.scientificconservation.comNational Instruments Corpwww.ni.com
Subscribe to Automation World's RSS Feeds for Columns & Departments
