Fieldbus itself is hardly new, but with big business benefits being missed, industry needs to wake up and reconsider its factory roots. Andrew Ward explains
Digital factory-level networking technologies are increasingly bringing business benefits to manufacturers partly because of the much more sophisticated real-time communication they allow with automation devices, and partly as a result of the greater flexibility they permit. There is great practical and financial value in being able to accommodate change, however it manifests itself.
Unfortunately, so-called ‘fieldbus’ plant/factory network technology remains confusing. For too many years it suffered the ugly side of competition, with procrastination, misinformation and the paraphernalia of vested interest preventing the essential standards and user acceptance. And with the arguments drawn out and apparently largely technical, although actually driven at least as much by commercial protectionism and entrenched dogma, the muck stuck. The result: users, did they but know it, got a raw deal. Although the fruits of standardisation are now largely there for the picking, with vendors by and large forced to open their fieldbus technologies so that all can build compatible devices, the legacy is still taking some shaking off.
Confusion, together with healthy scepticism, has led to a still painfully slow take-up of fieldbuses over their age old primitive analogue hard wiring alternatives, certainly in Europe. Another barrier has centred on users’ perception of price and technology. Ethernet, now the universal foundation for office networks, is far cheaper than industrial alternatives, but remains unsuitable for much of plant and factory control. As Heinz Janiec, group leader for process computers at Shell & DEA Oil GmbH, says: “Below the control system you need a way to operate things in a deterministic environment. If you introduce switch technologies you can make Ethernet deterministic, but switches are currently expensive.”
Ethernet’s deterministic behaviour, or lack of it, and concerns over power carrying, particularly in classified hazardous areas, are both currently being addressed by development and standardisation efforts. But in the meantime, for users considering fieldbus technology now for device-level networks – rather than information-level at the top, or wiring replacement ‘bitbus’ networks at the bottom – the realistic choice is between the various ‘flavours’ of standard Profibus and Foundation Fieldbus.
In fact, your choice of bus is best dictated by your needs and your industry. Consider whether or not there are devices available for hazardous areas, whether you need a bus certified to EN914 that can carry safety-orientated and standard communication on the same cable, or whether you need to send and receive parameter data without interrupting the flow of real-time sensor information. These will narrow your selection.
For example, “We have introduced Profibus DP at a jetty terminal because we have a lot of discrete operations there,” says Janiec. “Support for discrete operations is a big weakness of Foundation Fieldbus.” However, FF has advantages that make it a good choice for many process industry users. “Because of the distributed intelligence, even if the controller fails, as long as you still have power you still have operation – a sensor can talk to an actuator, or vice-versa,” says Janiec.
You’ve got it: the danger is you might end up with more than one type of bus. That isn’t anything to worry about, according to Mark Daniels, marketing manager for drives at controls vendor Rockwell Automation: “Look at what the control PCs will do for you, and what the software will do – that’s what’s important,” he advises. You may not like the prospect but he insists you can treat your environment as one complete control system, regardless of different devices being connected on different networks.
In any case, many automation device vendors now offer support for a wide range of protocols – to a greater or lesser extent. For example, Foxboro, part of Invensys, supports FF, Profibus, HART (the older Fisher-Rosemount approach involving digital signals superimposed on two-wire analogue leads), and proprietary FoxCom digital communications.
The bottom line is that whatever technology you go for, the benefits really can be considerable. David Humphrey, senior consultant with analyst ARC, puts it thus: “Installation costs – both materials and labour – are reduced dramatically compared to point-to-point harnesses, but there are other benefits, although they can be more difficult to measure.”
Multiple intelligent benefits
Most important, using smarter digital networks means you can get the benefits of so-called ‘smart’ digital plant devices – which provide valuable information in several respects. Especially applicable to process industry users, smart transmitters deliver potentially much greater signal accuracy and eliminate the old problems of scaling and out-of-range values that can bedevil 4-20mA analogue transmitters. And all that means better feedback loops, controls capable of running closer to the wire and improved quality maintenance, with the usual results of reduced operating costs and possibilities in terms of price differentials.
Smart devices also make problem-solving and general operations easier and more efficient. They can automatically run diagnostics so when you or your system read data you automatically know its credibility. What’s more, since modern plant and factory control systems have the intelligence to read the diagnostics, they can automatically take appropriate measures to mitigate the effects of device drift or failure.
Then again, the ability to remotely program devices and set their parameters has several benefits. There’s no need for someone to physically visit the device, itself a significant gain given issues of time and difficult access. Further, easier programming enables you to reconsider flexible manufacturing and smaller batch sizes from the ground up. Changing the acceleration of a conveyor system to accommodate a special bottle for a promotional range, for example, can be carried out by a controls engineer from an office. What’s more, the system tracks changes, so the next shift will understand what’s different, and can easily roll back.
The list goes on: there will also be savings on everything from initial installation and testing, to implementing change. As Mike Skidmore, director of consultancy Aims for Industry, says: “Within a discrete manufacturing facility, if a new process is developed – perhaps a hardware process controller – the ability to introduce that within one or two test cells relatively easily so the benefits can be seen, and then to roll it out throughout the plant, is much better than the disruption you’d have with traditional wiring.”
And at the back end, as Humphrey says: “You can save a lot of money through preventative or predictive maintenance. You increase the life of the machine, and get more from your assets over a longer time.” Smart devices in the factory not only tell you when their processes need attention, but also when they do. It’s a detail, but an important one – ideally, you want to maintain a valve before it fails, but preventative maintenance can cause downtime and waste money. A smart actuator on fieldbus tells you how many times it’s stroked since last maintenance. It can tell you exactly how far it has moved in response to a control command and schedule intelligently timed asset management.
OK, it’s convincing on paper, but hard returns on investment are difficult to calculate. Savings on wiring and installation are one thing to cost, making a fieldbus attractive for a new install, but there are usually inescapable higher purchase prices for smart devices.
But the biggest potential for benefit, and the most difficult to calculate, is in the more intelligent and flexible manufacturing that fieldbus allows. And now you need to go beyond total cost of ownership to serious strategic thinking. What if you could…
As ever there are also some words of caution, the first from Janiec: “The major difficulty is to get the right skills and training for the instrumentation engineers. It’s a completely different world,” he says. And Brian Holliday, communications manager with Siemens Automation and Drives, adds: “Manufacturers will overlook the cultural issues at their peril.”
But the up-side is that operators get more and better information, and as far as they are concerned, the means by which it’s delivered is transparent. A little training and re-skilling, and some slightly higher up-front costs seem very small prices to pay for better, more efficient, more flexible, less hide-bound, less wasteful manufacturing and manufacturing management.