Digital Diagnostics

Half of the respondents to a recent Automation World survey indicate that they do not use the smarts—the advanced diagnostics—available in their Hart and Foundation Fieldbus devices. The reasons: cost, complexity and training.

Manufacturing and production plant systems offer engineers endless opportunities for problem solving, both in system design and day-to-day operation. This is because no single element of a system remains perfect. Everything is dynamic, and a goodly percentage of the movement is in the wrong direction. This is where diagnostics come in. Problems have causes, and the aim of diagnosis is to identify causes. Nail the cause and you are well on the way to nailing the problem.

And this, in turn, is where the diagnostic capabilities built into smart Hart and Foundation Fieldbus instruments apply. Hart is a communication technology and standard promulgated by the Hart Communication Foundation (hartcomm.org), while Foundation Fieldbus is a network developed and promulgated by the Fieldbus Foundation (www.fieldbus.org). Every supplier of bus-system-ready instruments offers versions that are capable of focused and detailed diagnostics, and these diagnostic subsystems are eagerly harnessed by end users—or are they?

A recent Automation World survey indicates that many engineers do not use the diagnostics offered to them. The survey, less about statistics than opinions, revealed that of the 60 identifiable Hart and Foundation Fieldbus users, about half use the available digital diagnostics.

“Diagnostics” does need some definition. It is a market-positive word and like most market-positive words, manufacturers apply it to a variety of capabilities. “Smart diagnostics,” which is relevant here, goes beyond device- and device-transmitter health. It is more than simple feedback that a device is open or closed or pushed or pulled when receiving a control signal. Advanced diagnostics is the sophisticated feedback and monitoring that can be built into a device. In other words, for a valve we mean flow rates, pressure, temperature and the like, not simply that the actuator fired and the valve is open or closed.

Actual diagnostic capabilities can be a complex mix that is dependent on the device type and the manufacturer. The Hart Foundation’s Application Guide puts it about as succinctly as possible:

“Hart instruments can implement extensive, device-specific diagnostics. The amount and type of diagnostic information is determined by the manufacturer and varies with product and application. Diagnostic information can be accessed using the Hart Communication Protocol. Host applications using DD [device description] files can interpret and display diagnostic information. Applications not using DD technology may require product-specific software modules to interpret diagnostic information.”

Here is the key question: Why have virtually half of users opted out of using smart diagnostics? One common answer: lack of integration with other devices. Other reasons include:
 

  • the lack of trained personnel to implement and manage or interpret the diagnostics;
  • lack of input/output on legacy distributed control systems (DCS);
  • the need for additional software and systems; and
  • related costs.

Integration issues

There was a fair amount of top-down approaches implied in some of the answers. Among these: “No [we don’t use the diagnostics because] they are difficult to integrate into a broader network.” “No, [because we] need to connect to other standards.” Even the yea-sayers recognize issues around data silos: “In general, yes [we use diagnostics] although the diagnostics are difficult to move into databases and other control systems.”

ON THE WEB: Hart Communication in Action
Read why Wisconsin Energy Corp. (WE) won the Hart Innovative Achievement Award. WE integrated Hart into its maintenance procedures and device configuration.
Visit bit.ly/awnews_133

Broader networks, other standards, challenges moving data: It is heartening to see users embrace this kind of systems approach, one that says there is a need to look at several devices to see a broader picture than that which is available in a single instrument.

Unfortunately, the richer the diagnostic potential in each device, the more potential there is for complexity in the process of interconnecting them. As one respondent put it, “The large quantity of diagnostics in devices does create issues with what to use, and [affects] the cost/value ratio in the selection. In most plant applications, there is no need to use any diagnostics…. [There are] only a very small number of installation points where they really add value over the cost to implement, monitor and action the outcomes.”

The toughest aspect of engineering is simplifying—in this case, paring down to the essential few devices whose data should be mined in detail—and the larger the system, the longer will be the discussion around what to monitor fully. But, as the same respondent wrote, “The real point is, if you are not going to action them, then they are a wasted cost.”
Lack of trained personnel seems to be an important obstacle. We tend to forget that with the advent of every control breakthrough, more engineering complexity arrives as well. As one respondent put it, “No [we are not using diagnostics. Vendors] need to make it easier to get to the information.”  Another said, “[In] most cases the extra components and complexity add more issues then they solve. The information set-up is done wrong, [and] in the end the information that matters does not [rise] to the surface.” A third said simply, “No, we don’t have enough qualified personnel to do this kind of diagnostic.”

One respondent put it most succinctly: “No [we don’t use diagnostics] because the employees don’t have all the necessary skills to take advantage of the available information.”

New vs. legacy systems

As time passes, there is an ever-widening gap between long-established process plant sites and new ones. Computer and control technologies move forward all the time; not so, the welter of older components at older sites. Joe Shingara, marketing director for automation systems supplier NovaTech, points out, “To a large extent, customers with mature operations haven’t employed smart instrument diagnostics. Customers in upgraded or greenfield sites do use them. You don’t see people doing wholesale change-out of existing instrumentation at older plants—the costs are a major factor.”

Not surprising, then, are the survey comments around this issue: “The DCS systems we use don’t support Hart diagnostics.” “Device diagnostics mostly remain trapped in the devices, since we are just starting to move from 4-20 mA to include Hart.” “[We are not using diagnostics] at this time as we have not implemented a system. That’s partly because of the complexity of doing so, partly because we’re short staffed, and partly because of the cost of implementing a new program.”

A few of the nay-sayers indicated that, to use diagnostics, they would have to implement additional software or systems—and to do so would require more than they want to invest. As one respondent wrote, “Asset management apps are mostly offered as separate features, not easy for users to implement strategies at low cost.”

Other half: Eager users

Although the percentage of nay-sayers is high, we do have to remember that half of the respondents are ready, willing and able to take advantage of advanced diagnostics: “Sure, Fieldbus diagnostics are very useful in troubleshooting field issues,” said one. Said another, “Yes, we are getting valuable outcome from Hart, including calibration ([where we determine] precision, accuracy, sensitivity, damping time adjustment, digital technique etc). [The availability of] process alarm settings and diagnostic messages, and user friendliness compelled us to use [the diagnostics.]”

A couple of respondents cited the positive side of diagnostics during commissioning, where the ability to look into the operating parameters of devices is valuable. As one respondent wrote, “…yes, we get all [possible] value from it. [We can speed up processes] and process dynamics can be fine tuned.”

One gets the feeling that the passage of time will resolve much of the reluctance to use smart diagnostics—at least when the issue is that many existing devices came on board long before smart instruments were born. Still, process equipment and devices have to be swapped out eventually, and while this often begins with a patchwork approach, in a healthy company, patching eventually gives way to a fresh, system-wide view of what has to be done. In the best of worlds, a system view evaluates all available tools and capabilities.

A rich source of insight flowed from the survey question, “What could suppliers of instrumentation or asset management software do to help?” Inevitably, among the most commonly stated or implied solutions was, “Reduce the price.”
Time is on the end-user’s side on this aspect. Digital equipment continues to deliver ever-greater power in ever smaller packages at lower costs. Also, many sensor technologies become less costly as once-exotic approaches become commonplace.

Economics aside, survey respondents offered other insights worth contemplating. Many asked for stronger process and operator views. As one respondent wrote, “Make it more targeted at applications rather than the equipment.” In other words, replace a valve’s diagnostic message that reads, “0.5 l / min. flow rate” with one that says, “Flow rate low. Tank depletion in 45 minutes.”

How might this be done? Some responses: “Improve the reliability and pinpoint exact issues with various diagnostic errors.” “[Provide] user friendly device alerts with an overview page highlighting the event and possible causes, without having to drill deep into the diagnostics.” “Improve the reliability and pinpoint exact issues with various diagnostic errors.”

Unfortunately, while the concept is simple, the software development implications are huge, simply because every process is in some way unique. Focusing on the instrument is far easier than delivering a message such that it relates to one specific process. A single number easily captures, for example, flow rate, but capturing the implications of that number is far more difficult. Still, as one person put it, there should be ways to “make it easier to set up or provide free or low cost assistance in setting it up and making it work.” Another said, “More information is key. Do not assume that the reader knows anything about setting up a device. Diagrams help. Step by step for different applications.”

Finally, there were several responses calling for greater standardization and its corollary, less segregation of protocols and methods. Several suggested Ethernet network connectivity. Several asked for tighter integration into DCS and supervisory (SCADA) systems.

In the final analysis, what our survey illustrates is the glacial rate at which technologies penetrate process industries. In some respects, commissioning a new process site inevitably becomes the first step in setting that site’s systems in stone. Although everything is fluid during design—and may be quite fluid for the first few months (or years)—solidification soon makes change difficult. The investment both of time and money is huge, and the cost of change tends to insulate everything from the continuously changing technical scene at suppliers.

That said, diagnostics are almost certain to win out. As one respondent put it, “They can help detect failures before they occur. We can trend the data before anything impacts the process.” Total up the costs of unplanned downtime and the return on investment, and use of available smart diagnostics becomes increasingly attractive.

Dave Gehman is a Contributing Editor for Automation World.

Fieldbus Foundation (www.fieldbus.org)
Hart Communication Foundation (hartcomm.org)

More in Networks