Real time networks: think strategically

UK companies are lagging behind the US in adopting digital ‘fieldbus’ (sensor/device communications) technology. So says Kevin Prouty, senior analyst for AMR Research. And this means they are missing out on the associated business benefits. To make matters worse, the long-running debate (more than 15 years) continues between the various industrial protocols: Fieldbus Foundation, Profibus, Interbus, DeviceNet, ControlNet, WorldFIP, and emerging standards like CANbus and Ethernet. Just as progress appears to be on the immediate horizon, another standard emerges and we are back at square one again. Little wonder end-users are confused. The term ‘fieldbus’ refers to any type of low level digital industrial network, and is an enabling technology that brings its users a host of benefits, albeit dependent on the plant’s current technological state. For starters, savings should come from using a single network cable rather than lots of hard-wiring and I/O hardware. As Dr Richard Piggin, senior analyst at ARC, points out: “The benefits most frequently cited by users of fieldbus have been flexibility, design, speed of commissioning, diagnostics and maintenance. The exchange of digital information is also important, providing more accurate data capture than analogue systems.” There’s more though. Adopting fieldbus means lower cost, PC-based predictive maintenance and overall plant asset management become possible. And having well-maintained machines and equipment means an increase in machine (or process) availability, thus improving plant productivity and quality of product. This equates to higher profits and fewer returns. And then there’s routine condition monitoring. The argument is, it becomes easier to do with fieldbus. It’s easy to monitor all the critical parameters in a process (or machine). Richard Sturt, UK business manager automation and control at Rockwell, comments: “With digital fieldbus, for example, you can measure the torque or thermal loading on a servo motor that may be a key part of a high speed packaging machine. If there’s a problem that causes the motor to overheat, this would be picked up by parameters monitored over the network. You can alert the operators, in real time, before the system actually stops. “There are over 200 parameters in our ‘simple’ servo drives – you simply don’t get that kind of information without [digital] networks,” he adds. But if the benefits of digital networking are so clear, why has fieldbus adoption been so lacklustre in the UK? Part of the answer lies in the well-known cultural differences between the IT department and control engineers, but the real issue here stems from who actually makes the buying decision. The fact is that within most UK manufacturing firms the decision to buy is usually a very fragmented one, made by the plant or process engineer. Analyst AMR’s Prouty puts it another way: “It’s like five different engineers in a company, working on five different projects, each purchasing five different fieldbuses – without any corporate or strategic view of the plant. “In the US, we’re already seeing a huge shift from this type of fragmented approach, to a corporate IT one. In other words, when an engineer buys a PLC, he now needs to consider the company’s overall networking strategy: he cannot simply make an isolated decision to buy,” he claims. It’s not all doom and gloom for the UK though. Some manufacturers are making good strides. Federal-Mogul Powertrain Systems, based in Bridgewater, manufactures pistons for new and rebuilt heavy-duty engines. With approximately 450 employees, and an annual turnover of around £40 million, the site is part of a global group which encompasses design engineering, component manufacturing and complete systems delivery. The company has just finished retrofitting one of its manufacturing cells. John Elkins, maintenance manager at the plant, recalls: “One of our casting cells needed a complete rethink. There are two casting machines and a robot in the cell, nothing complicated, but everything [sensors, relays, etc] was hard-wired. The problem was that we had frequent spillage of molten metal from the casting machines, which kept damaging the wiring looms. We needed a fast solution, since re-wiring was taking up to four days in some cases.” He adds: “After considering Profibus, we opted for a programmable safety system (PSS) from Pilz Automation. It runs on the CAN fieldbus – an emerging technology from Bosch, used on car wiring systems. “We are now in the process of commissioning and testing the cell,” he adds. “And now, if we get a safety feature that doesn’t work, we simply reprogram using the PSS. It takes three or four minutes instead of three or four days of hard-wiring.” Elkins reckons installation costs were around £8,000 (including training). “That’s only about £1,000 more than hard-wiring would have cost,” he explains. “But for the complete cell, we’ve saved around four weeks of time. Just think what we could save if we used a similar approach across all our manufacturing cells.” And he’s right. The statement sums up the current state of many UK manufacturing plants. The engineer, who is often close to the problem, makes an isolated decision to go for best-of-breed technology, integrated. And the result is only modest gains. As Piggin explains: “The real benefits only start to emerge when companies plan their networks on a strategic level throughout the whole plant, or across several sites.” As for the foreseeable future, Piggin predicts: “People [users] have been sitting on the fence naively waiting for a unified standard. Ethernet certainly has its place, although not at the device level. But expect to see a thinning out in the control layer. The plant floor automation hierarchy will collapse as more devices have Ethernet connectivity.” But will Ethernet be used for safety-related applications? According to Piggin it already is being used. “It’s available now for critical process control. Many of the existing protocols are being developed with an additional safety layer, whilst the same protocols are also being developed for Ethernet. A combination of the two is inevitable.” Whatever happens, over the next few years, it will be interesting to see which of the three technologies: Industrial Ethernet, (the several) fieldbuses, or conventional wiring, the majority choose to go with.