Control News from Europe December 2007

Dec. 17, 2007
Andrew Bond reports on process controls

By Andrew Bond, Industrial Automation Insider

Honeywell Does Austria

Salzburg, for the second time this year and the third in 20 months. This time it was the EMEA Honeywell Users’ Group or HUG Conference. Despite a reported record attendance this seemed a slightly more subdued event than either last year in Seville or 2005 in Prague. But if this year’s event seemed a little short on major announcements, there were plenty of less headline-grabbing developments which, taken together, could be said to add up to some important shifts in emphasis. Arguably the most significant of these is a renewed interest in, one might almost argue a rediscovery of, field devices or, as Honeywell likes to put it, Level 1 in the Purdue model.

As one Honeywell executive put it to INSIDER in a subsequent conversation, HPS has had to make its mind up whether it is going to be in the field device business as a serious player or whether it should simply get out of it altogether. The decision to take it seriously, he suggested, had reenergized a section of the company which had hitherto been, or at least felt it had been, treated as a poor relation.

One of the key factors in this change in emphasis has been the recent acquisition of Enraf, and it was former Enraf vice president of international sales and marketing Tony Tielen who was tasked not just with explaining the Enraf acquisition, but with presenting the new strategy.

However, it is also clear that wireless has proved a catalyst, raising the profile of field devices as vehicles for delivering added value. No surprise, therefore that, in Honeywell speak, we’re no longer talking about field devices, but ‘field solutions.’

“Look for Honeywell to make more acquisitions in this space,” said EMEA vice president and general manager Paul Orzeske in what would have been his opening presentation, had his flight from Chicago arrived on time. What kind of acquisitions? “We’re changing the business model about how we deliver solutions in Level 1. We’re migrating from a pure technology to a solutions provider and we’re looking for acquisitions which bring solutions.”

Custody Transfer

That seems to be exactly what Enraf has delivered; for example, not just specific technology for,
level gauging, but complete solutions for tank inventory management, blending and additive injection, loading systems and custody transfer. According to Tielen, the aim is to enhance the field device portfolio to include not just all the main process measurements but, more importantly, to add differentiating technologies such as on corrosion measurement, wireless and weights and measures. Combining these technologies with domain expertise will enable the company “to excel in the deployment of next generation instrumentation.”

That will in some cases mean acquisitions; in others, collaborations such as that with Krohne to develop the recently announced VersaFlow range of mag flow, Coriolis and ultrasonic flow transmitters. These, Tielen emphasized, aren’t just rebadged devices but joint developments, with Honeywell in particular adding its wireless expertise. The result is a rapidly developing know-how in level and flow, with a particular emphasis on custody transfer.

No one is seriously expecting Honeywell to be challenging the Rosemounts, Yokogawas or even
Endress+Hausers of this world as field device vendors in the near future, but there is nevertheless a clear determination to make an impact at Level 1, something which Orzeske believes is essential for a Main Automation Contractor (MAC).

But Level 1 is certainly not the only area where Honeywell seems to have shifted its stance somewhat. Another is the still highly contentious area of Safety Instrumented Systems (SIS) where Honeywell has spent the last couple of years perched on the fence between the advocates on the one hand of full integration of BPCS and SIS and, on the other, of total separation.

Closer integration

Now, while there’s no lurch in one direction or the other, there are signs of a discrete shuffle toward the integration end of the spectrum. Inevitably Honeywell argues that it’s always been able to provide a degree of integration, and it still stops well short of an ABB or Yokogawa style solution based on a common hardware platform, but the next release of Experion, R310, due in early 2008, will not only include new hardware for Safety Manager, but also a common HMI across BPCS, SIS and Fire & Gas. EMEA director of sales support, Jean-Marie Alliet, stresses that what we’re talking about here is “Operational integration which avoids common cause faults.”

That common HMI is also a major theme in the drive to improve operator effectiveness, based on Honeywell’s involvement in the Abnormal Situation Management (ASM) consortium. While much of that is already familiar, what’s new is data showing, for example, a 38% improvement in the ability of an operator to recognize the development of an abnormal situation before the first alarm and a 35% to 48% improvement in the speed of response. This work is now extending into a key area for improved safety through the introduction of automated procedures for infrequently performed operations, such as start up and shut down. With procedural operations implicated in 30% of all reported incidents, the aim is to use automation drastically to reduce, if not eliminate, human variation and the dependence on operator experience. It’s an approach that echoes the view of exida co-founder Rainer Faller who, in a keynote safety address, suggested that the continuous process world has much to learn from the batch world when it comes to the safe execution of procedural operations.

Watershed

And what of wireless? Jack Bolick, speaking via video, described it as a “watershed event” and Jean-Marie Alliet predicted that it would usher in “a technology step change like the DCS 30 years ago.” Honeywell, claimed Tony Tielen, was the number one provider with the largest installed base and was “endorsing and driving SP100.” Missing from any of these pronouncements, not surprisingly, was any mention of the standard that dared not speak its name, at least in those surroundings. Indeed OneWireless senior product manager Yu-Gene Chen  managed to get through an entire one-hour wireless presentation without using the word WirelessHART once. So where does Honeywell now stand on the issue, and why did they oppose the release of HART 7.0 and with it WirelessHART in September?

According to Harsh Chitale, vice president for strategy and global marketing, the issue is all about interoperability and flexibility and the management of the available spectrum. SP100, he argues, offers the more elegant and complete solution with a single network and, crucially, a single security solution for all protocols, including HART. Moreover, he claims, it provides the most effective means of ensuring that competing applications can coexist without interference either from each other or from applications on adjacent sites. “There’s only one spectrum to share. There has to be a holistic approach if users want to use wireless.”

With rather more confidence in the ability of the ISA100 committee to deliver than some would consider justified, he anticipates an ISA standard by the second half of 2008, which he suggests will be only a matter of a month or two after test and certification procedures are in place for WirelessHART. So, the argument goes, why rush ahead with a solution which only addresses device networking, when a solution for the entire gamut of applications will be available just months later? Or, to put it rather more colourfully,“Why would you want a pager when you can have a Blackberry?”

Interoperable

But the problem for Honeywell now is that we are where we are. So are they going to support WirelessHART now that it is a fact? Jean-Marie Alliet says “We will support it if we need to, that is if that is what customers are asking for.” For Chitale, however, the issue is how WirelessHART is accommodated within ISA 100. In the first release, the aim will be to make the two interoperable, but the ultimate aim should be to achieve seamless integration with a common communications stack. That’s possible, he says, but it will need movement on both sides, and at present he doesn’t see the necessary flexibility from the WirelessHART camp.

Given that they find themselves in a minority of one on WirelessHART, the Honeywell team seems remarkably relaxed. Will further delay hurt them? “It’s not a problem because for us sensors are only a part of it,” says Chitale. And what happens if there is no agreement? “It’s not going to be like the fieldbus wars. This time there is a clear economic benefit for the users and they will drive it.”

With all the arguments over standards, it’s easy to forget that there are some real divisions of opinion over the technology itself. One of the defining statements about the technology on which WirelessHART is based is the comment from BP’s David Lafferty that “It just works.”

Not according to Honeywell’s Yu-Gene Chen. “It doesn’t just work,” he says. “We have a self forming, self healing network too but there are still problems,” said Chen, highlighting what he believes is the fundamental importance of site surveys. “Other vendors say you just have to put in another node but if you do that you create other problems. Every time you make a hop, you cut throughput by half so, pretty soon, you have an unusable network.” Can’t think who those other vendors he’s referring to could be, but there might just be a response when Emerson brings its latest wireless offerings to Europe at the end of the month.

APV sale rekindles bid rumors at Invensys

Both GE Fanuc Intelligent Platforms and Rockwell Automation are trying to present themselves as serious contenders for a share of the mainstream process automation market. Jim Pinto, of course, still has Rockwell down as an acquisition target for ABB, a fate which, he argues, can only be avoided by itself making a major acquisition, by which he means something an order of magnitude larger than any of its 2007 deals to date. This year saw Rockwell for the first time figuring in ARC’s global DCS market share rankings but it’s difficult to see how either it or GE Fanuc could make a significant impact on the rankings without buying into an existing contender’s installed base.

Who to acquire? One rumor current in the US, we’re told, has GE in the process of purchasing a company valued between $500m and $1b with batch, hybrid and DCS capabilities. That, so the argument runs, can only be the ‘hardware’ parts of Invensys - Foxboro, Triconex, Eurotherm, APV – whose disposal would leave IPS free to concentrate on software and services.

SPX

Unfortunately Invensys has somewhat spoilt this one by selling APV to SPX Corporation for £250m ($518) cash. APV has been a thorn in the Invensys side for years and, despite repeated efforts to improve its performance, has remained the group’s lowest margin unit. £70m of the proceeds will go to reducing the shortfall in the pension fund while the remainder will be used to pay down debt.

The news pushed the shares up nearly 6% on the day to close at 327p, and they’ve since been boosted further by speculation for the whole group at some 450p. Wiser heads, however, suggest that a serious bid is unlikely to materialize until Invensys has disposed of its other black sheep, its Controls division.

Departures

Meanwhile the word on the Foxboro street is of further turmoil within Invensys Process Systems (IPS) following Paulett Eberhart’s decision to move the IPS HQ to Dallas. Most prominent among the casualties are Measurements & Instruments business unit general manager Ken Brown and Wonderware president Mike Bradley. Brown, it will be recalled, was widely tipped for the top job after Mike Caliel’s departure in June 2006 and even took over the role until Eberhart’s appointment in January 2007. Meanwhile Bradley’s departure comes only weeks after the launch of InTouch 10.0 and System Platform 3.0, which he himself described as the most important announcements in Wonderware’s 20- year history. Bradley is, we understand, to be succeeded by Sudipta Bhattacharya, who joined from SAP only last August as Invensys’ chief software solutions officer, reporting to Bradley.

Departures at less elevated, but equally, if not more, painful levels include all but one of the IPS Marketing Communications team, as MarComms follows the HQ team to Dallas.

The cliché now standing at platform . . .

Automation vendors have been vying with each other for some time now to incorporate the magic word ‘platform’ into their hardware and software product descriptions—Rockwell, for example uses the term to describe both FactoryTalk and Logix—and, even more innovatively, into actual product names such as Wonderware System Platform. Now, however, GE Fanuc Automation has raised the stakes even higher by renaming the whole company to incorporate it and, even more daringly, coupling it with another overworked adjective to create GE Fanuc Intelligent Platforms. However, while the old name actually gave some indication of what the company was about, the new moniker leaves all but the most intimately initiated guessing.

Does the fact that the new branding was slipped out largely unannounced suggest that even within the hallowed halls of GE, there are those who are just the tiniest bit embarrassed? If they are, they’re just going to have to get used to it because the change of name presages the launch early in the new year of GE Fanuc’s own version of the platform concept under the entirely original name, Proficy Platform. Previewed by CEO Maryrose Sylvester at the company’s Discover 2007user conference in St Louis, Missouri, last month, this new platform and its companion, Proficy Workflow, will exploit the new Microsoft technologies to achieve what she called the ‘Googleization’ of manufacturing software. Hmmm!

Wireless

Meanwhile, having got the serious matter of what the company is actually called out of the way, GE Fanuc has become the latest vendor to toss its hat into the increasingly crowded wireless ring by announcing TranSphere Wireless, a range of solutions embracing extended range IP networking, wireless IP/Ethernet connectivity, Ethernet and serial communications radio modem offerings and remote I/O for analog and discrete I/O signals. “Wireless communication is a major trend that is gathering strength in the automation arena,” explained Controllers product manager Bill Black who then damaged his own credibility somewhat by asserting that, in entering the market months if not years after some of its largest rivals, “GE Fanuc is at the forefront of this wave as we now offer our customers the ability to integrate wireless communications into their automation environments.”

TranSphere has more than a whiff of the OneWireless about it, being promoted as a means of reducing the integration, configuration and support costs incurred by what are referred to as ‘multi-box’ solutions. It is said to provide users with the freedom to align with their individual distance and I/O requirements, and it’s also being offered for off-site applications such as tank farms, pumping stations and treatment plants. The actual devices support multiple users connecting to multiple applications via multiple protocols on the same unit or the same network, and provide for multiple layers of protection, including 128-bit data encryption, two-way authentication, and dynamic key rotation. They’re also offered in 2.4GHz and 900MHz versions to cover the market requirements both inside and outside North America.

GE reckons that TranSphere offers the longest range industrial products in its class and that it supports open standards, although it doesn’t actually say which ones. To that extent it really is truly unique—a wireless offering announced without any mention of ISA 100 or WirelessHART!

CPS sounds the death knell for MES

The latest attempt by ARC to rewrite our automation vocabulary—remember Process Automation System or PAS for DCS and Programmable Automation Controller or PAC for PLC—comes with the launch of its Collaborative Production System or CPS model which had its first public outing at GE Fanuc Intelligent Platform’s Discover 2007 user conference in St Louis last month.

Developed by the ARC team of Larry O’Brien, Tom Fiske, and Craig Resnick, CPS “articulates the collapsing of the separate domains of automation, operations management and plant engineering and design, breaking down the barriers to information and forming a framework for operational excellence in the manufacturing industries” which, we gather from reports of Resnick’s presentation in St. Louis, is ARC speak for replacing MES.

Simpler information flow

And for once you can see why. MES, which is in any case in its second incarnation after a less than satisfactory and mercifully short life in the ’80s, has always meant pretty much whatever any vendor has wanted it to mean. It’s less a coherent concept than a rag bag of more or less independent applications, such as scheduling, inventory, OEE and quality.

ARC describes the new model as combining its Collaborative Process Automation Systems (CPAS) and Operations Management models, outlining both systems infrastructure and the requirements, functions, people and processes needed to achieve operational excellence. “Collaborative Production Systems prevent assets from being niche islands of information, and ensure that all assets are delivering their maximum return on investment to the manufacturers and their shareholders,” explains Resnick. “Collaborative Production Systems also eliminate manufacturers’ internal barriers that may exist between, for example, plant floor and IT personnel.”

Collaboration

The model also provides direction to manufacturers on how to apply the principles of collaboration to ensure that all of their production systems share relevant information with each other. The architecture reflects the increasingly distributed nature of applications, creating opportunities to improve performance in such areas as asset and equipment reliability, automation and information context, knowledge worker enablement, plant performance intelligence and common actionable KPIs. CPS also acknowledges the challenges presented to manufacturing companies by the “flattening” of the world as they respond by shifting focus to flexible customer-centric manufacturing to deal with demand fluctuations. “ARC is a strong advocate of manufacturers adopting solutions that subscribe to the CPS model which will go a long way to removing the barriers between Operations Management and Automation Control,” said Resnick.

By presenting CPS at its user conference, ARC appeared to be endorsing GE Fanuc’s approach, an endorsement which GE wasn’t slow to exploit. “We share the vision for breaking down the barriers associated with the traditional ‘layered’ production environment with thought leaders such as ARC,” said Control Systems vice president Bill Estep. “The CPS model and our solutions are closely aligned to deliver value to manufacturers globally. Dissolving the walls and eliminating boundaries that our customers have faced between Operations Management and Automation Systems is exactly what we have been focusing on with our solutions, and the ARC CPS model clearly articulates the next generation model for those seeking operational excellence.”

Whether any of that increases the chances of us all stopping talking about MES and latching on to CPS as the new must-have acronym remains to be seen, but previous experience is not encouraging. After all it’s the very vagueness of MES which makes it so useful to vendors, allowing them to claim it or deny it as expediency dictates.

Does gateway commit ISA to a WirelessHART future?

ISA and the HART Communications Foundation (HCF) seem to be making up for lost time in their attempts to achieve a single wireless standard. Following their initial decision to collaborate on seeking ways to incorporate the newly released WirelessHART protocol into the ISA 100 ‘Wireless Systems for Automation’ standard, they held a series of meetings at last month’s ISA Expo event in Houston, which appears to have charted a possible way forward.

That initial decision to collaborate was prompted, at least in part, by Honeywell Process Solutions president Jack Bolick’s last-minute attempt to hold up release of the HART 7 protocol on the grounds that it would lead to the emergence of duplicate standards and hence confusion among users.

Endorsement

It looks very much as if similarly strongly expressed views from other leading vendors have been responsible for maintaining the momentum. For example, ABB Instrumentation technology vice president Sean Keeping wanted to avoid a costly repeat of the fieldbus saga which has led to ABB having to incorporate support for Profibus, Foundation fieldbus and HART into its field devices, tools and host systems. “This duplicated design effort has been costly for ABB and other suppliers, and has made life confusing for users. We would strongly advocate avoidance of a ‘multiple standard’ path for wireless sensor communication,” he said, to which end “. . . ABB’s recommendation is that SP100 incorporate the WirelessHART specification as its solution at the instrument level for PV Monitoring, Asset Performance Management and eventually control.”

His view was echoed by Siemens Sensors and Communication divisional president Hans-Georg Kumpfmüller when he said that recent events showed that “it is the wish of the industry to get more harmonized solutions, bridging the different standards existing on the market.”

Emerson chief strategic officer Peter Zornio and Endress+Hauser corporate director for Projects and Solutions Frank Hils added their voices in support, with all four companies indicating their intention of having WirelessHART-compliant products available during 2008.

According to ISA, the outcome of the Houston meetings was that the ISA 100 committee would “attempt to accommodate the HART 7 wireless protocol in release 1 of the ISA100.11a standard through a dual-gateway architecture, followed by a potentially more integrated approach in release 2.” ISA100 co-chair Wayne Manges of Oak Ridge National Laboratory said that the approach had the support of end users, suppliers, HCF representatives and the ISA100 members on the evaluation committee, who all saw it as “the most viable alternative to address the needs of the end-user community, both in the short and long term,” while his fellow co-chair, Pat Schweitzer of Exxon Mobil, explained that the joint Analysis Team was evaluating how
WirelessHART could be incorporated into the ISA100.11a standard without compromising the original ISA100 objectives. “The most important part of that evaluation is the obligation to continue our commitment to the end user, and we’re confident that our final decisions will accomplish that goal,” he said.

To that end the committee is to consider a number of options including dual-stack end devices for integration at the device level, tunnelling with pack at the device level and unpack at a higher level and future integration at the MAC, DLL and NET/TRAN levels.

All of which seems to be slightly at variance with the view expressed by Honeywell global wireless director Jeff Becker, who was reported as saying that Honeywell’s understanding was that the committee had agreed on “a dual-network, single-gateway compromise in Release 1, moving to complete interchangeability and a single network in Release 2.”

The difference may be more than just one of semantics. A Release 2 which embodied a single network which was not WirelessHART or at least WirelessHART-compatible would potentially leave stranded users who had invested in WirelessHART devices in the interim, which could be three years or more while, on the other hand, confirmation of WirelessHART as the only protocol for field device communication would appear to leave Honeywell’s own solution out in the cold.

Eighteen-month window

Actual ratification of Release 1 of ISA 100 is not expected before the end of 2008 which makes it unlikely that compliant products will be available before mid 2009. That would appear to give vendors of WirelessHART products a clear 18-month plus window in which to establish WirelessHART as the de facto standard, with the added validation of their being compliant with ISA 100 Release 1 thereafter. Given that a further year or more will have elapsed by the time Release 2 is complete, it’s difficult to see how ISA could contemplate a situation in which a significant installed base of compliant devices are effectively rendered non-compliant by a further release of the same standard. In effect, by conceding the case in Release 1, ISA100 seems to be committing itself to adapting its architecture to accommodate WirelessHART indefinitely.