No prizes for guessing what the press had come to hear about at Honeywell's EMEA (Europe, Middle East and Africa) User Group or HUG Conference in Seville, Spain in mid November. Sure, during the first morning's general sessions they showed polite interest in EMEA vp and general manager Paul Orzeske's review of Honeywell's current position on his patch and in senior technology consultant Jean-Marie Alliet's typically breathless and all embracing technology update.

However, it was the press wireless roundtable that followed, with contributions from ARC's Harry Forbes, 3e Technologies International vp John Jacobs - he's providing Honeywell with the means to overcome potential users' concerns over security - and Honeywell ACS Chief Technology Officer Dan Sheflin, that promised what the hacks had needed to convince their editors and publishers that four days in the Andalucian sunshine were not just desirable but essential. And that held true not just for European and South African colleagues but, rather more surprisingly, for the US contingent who, one must assume, had heard most if it at the North American HUG event in Phoenix back in June.

Don't mention the E word
According to Dan Sheflin, in any discussion of wireless technology in the process industries, "The elephant in the room is battery life." That may be true in general, but the elephant in that specific room that particular afternoon was undoubtedly called Emerson. And, while nobody but nobody actually dared to mention the E word, the entire purpose of the roundtable was not just to differentiate the Honeywell from the Emerson approach, but to demonstrate that it is incontrovertibly superior and a shoe in for the forthcoming SP100 standard.

Quite how successful it was in achieving the latter two objectives we'll leave for others to judge. For ourselves, however, we can say that they were successful in clarifying the essential differences between both the two companies' philosophies and the technologies. First the philosophy. Honeywell argues that wireless isn't just about data acquisition and control, although it certainly is about both of these rather than just data acquisition, but it's also about a host of other issues including mobile computing, asset management, security in the broader physical as well as the narrowly cyber sense and health and safety.

Overlapping technologies That could be seen as a response to Harry Forbes' major point, that we're not just talking about one or even one family of technologies here but about a host of overlapping technologies, the boundaries between which are already blurred and becoming more so. Different technologies are competing for individual applications and different deployed solutions are going to have to coexist within the same space, with perhaps four or five different networks based on different technologies by no means unusual. So this isn't the neat and tidy world envisaged by the standards aficionados and the critical capability will be to manage the wireless infrastructure at a time when the technologies are evolving very rapidly and the standards are having to evolve to keep pace. Which, incidentally, seems to beg the question of whether the whole standardization issue, at least as it's currently being addressed, isn't a huge red herring,

It's this all embracing approach which leads to Honeywell's concept of a 'wireless cloud' which blankets an entire facility, providing the infrastructure for the full gamut of wireless applications. Unfortunately, while that may be a vivid simile, it's one that rapidly generates confusion when it's translated into technological terms and into schematic diagrams. That confusion is compounded when different people are using the same terminology to describe different, if similar, architectures. And it's redoubled when those sensitive to criticism produce obfuscating or even perhaps deliberately misleading descriptive material. That's what's at the bottom of the argument that's developed over whether both the Emerson and the Honeywell architectures can fairly be described, as each claims, as self-healing mesh networks. Can they? We'll, yes and no. Quite simply we're not comparing apples with apples so it's hardly surprising if there's a certain amount of room for accidental or, more likely, deliberate misunderstanding. Does it matter? Probably not. What's more important at this stage is whether the different approaches work. On the strength of the demonstrations given in the exhibitions at the respective user group conferences, both do. More convincingly both can point to successful pilot installations in real world situations which Honeywell can supplement with some impressive statistics, even if some of them may not be strictly relevant: 300 wireless patents, 35 million wireless sensors installed, although the majority outside the process automation arena; and, more directly relevant, installations of the XYR 5000 wireless transmitter at more than 300 customer sites since its introduction in 2004, with shipments currently growing at 110% year on year.

Proven track record
It's those sorts of numbers which allow Honeywell to point to an existing track record in wireless and hence to imply that, by comparison, all of its competitors are Johnny-come-latelys to this particular market. But while they and others want us all to look at the wider picture, Emerson thus far has succeeded in getting us to focus on the specific issue of process measurement and it's that challenge to which Honeywell and the other players are being forced to respond. So what is the Honeywell architecture and how does it differ from Emerson's?

As we already know (INSIDER, October 2006, page 1 and October 2005, page 9) , Emerson's approach is based on a self-healing mesh network of wireless enabled field devices whose reliability improves as the number of devices increases, allowing signals to route themselves around obstacles and to find a new path if an individual device fails. There's no doubt that it's attractively simple in concept but, if Dan Sheflin is to be believed, there is a snag. He says that Honeywell tried this approach themselves and eventually abandoned it because, as we've already heard, of the battery life issue. Emerson has talked of other methods of powering the remote devices through various energy scavenging techniques but, as things stand at present, its devices are pretty much dependent on their batteries both to perform their primary function and to communicate over the network.

Battery life They've argued that battery life will be acceptable but Sheflin says the problem is that in large networks it won't be predictable; what is predictable, he argues, is that the larger the network becomes, the greater will be the traffic through the individual nodes and hence the shorter the battery life. That, he suggests, means that you'll be faced with the twin problems of unpredictable performance from the individual devices and an additional maintenance overhead. Technicians will be having to go out into the field to replace batteries when reducing the need for them to go into potentially dangerous situations was supposed to be one of the key advantages of the technology in the first place.

So Honeywell start from the standpoint that the nodes in the mesh network have got to be powered in order to ensure that network performance is reliable and predictable. These powered 'iNodes' thus provide the communications infrastructure not just for process measurement and control but for all the other services that Honeywell reckons you're going to want to handle wirelessly. The field devices themselves, on the other hand, don't participate in the mesh but communicate wirelessly simply with their nearest iNode. Despite what you've heard, according to Sheflin that doesn't require a line of site between the device and the iNode, although the latter will be elevated to get round the problem of tanks and vessels blocking the signal. Redundancy will be provided by allowing devices to communicate simultaneously with more than one iNode.

Crucially, says Sheflin, this means that the batteries in each device only have to provide the power for that device's own communication with the iNode which means that their life can be accurately predicted, based on the ambient conditions and measurement frequency. At 0C and a 1second measurement interval, he's predicting a battery life of 4.9 years, falling off to 3.3 years at 72C. That, he says, makes it suitable both for alarm and event monitoring functions and, potentially, for control, given that the network, unlike those of its competitors, supports latency control. At measurement intervals of 5 seconds and above, and at all but the highest ambients, he predicts a full 10 year battery life. By contrast, he claims, unspecified 'competitive' technologies - can't think who he's talking about - will achieve three to five years at best and substantially less in larger networks.

Might the need for power for the iNodes reduce the attraction of the Honeywell architecture? Sheflin thinks not, given that they'll usually be installed in existing plants at points where power is available, with the ability to locate the field devices up to 300m away. "iNodes are high power radios and very robust," he added. That, it seems to us, still begs the question of why, if they've got to have a power cable, they shouldn't have a signal cable as well. Perhaps that's the opening that Yokogawa's Uchida San sees for fibre optic technology.

Whether the Honeywell architecture constitutes a true self-healing mesh is one of those semantic, indeed almost religious, debates into which the process automation industry occasionally gets diverted. What is far more interesting is whether it and the Emerson solution can be reconciled within a single wireless standard. As things stand at present that looks distinctly unlikely although it wouldn't be too difficult to imagine Emerson networks acting as satellites to Honeywell's iNodes.

Users in the driving seat In any case Harry Forbes doesn't seem to think it matters too much. This isn't fieldbus revisited, he says, because this time the users are in the driving seat. They don't have to be convinced of the benefits; they already think that they're so compelling that they far outweigh the risk of backing the wrong horse. In any case, the penalties for doing so are largely mitigated by the fact that Honeywell, like Emerson, will be pledging to migrate its users to the eventual standards when it starts shipping its solution in the second quarter of next year. Before that, however, Emerson will be hoping to have the last word as it launches its own wireless offerings in Europe in the second week of January.

(IAI - December 2006)

Yokogawa makes haste slowly on wireless . . .

. . . while standards debate makes more fog than cloud With the decision time looming in the wireless standards committees it's perhaps no surprise that the fur is beginning to fly between the various contenders. Thus our report of last month's announcements from Emerson (INSIDER, October 2006, page 1) prompted the comment from a leading competitor that "The cyber- hackers and the Department of Homeland Security are going to love the Emerson wireless solution!" More specific criticism has been voiced of Honeywell's claim, voiced both by Process Solutions president Jack Bolick and business director David Kaufman, that it's technology, often described as a 'wireless cloud', is in fact based on wireless mesh technology. Unable to resist the temptation, "More like a wireless fog" was how one competitor described it, suggesting that because the Honeywell technology relies on multiple gateways and line of site paths between the devices and the gateways, it can't be a true mesh network. That's because the key characteristic of a mesh is its ability to 'self heal' in the event of a node being lost and its inherently increasing reliability as nodes are added, say the critics. Does the Honeywell technology do that? No doubt they'll be letting us know in due course and answering the philosophical question, "When is a mesh not a mesh?" Fieldbus dèjá vu Meanwhile Jim Pinto, never one to hold himself aloof when there's the chance to join in a good scrap, has added his own world weary, seen-it-all-before contribution. Quoting Gary Mintchell of Automation World he describes a standardization process which has become a confrontation SP-100 = SP-50 x 2 (smile)," he wrote in the latest edition of his Connections for Growth & Success eNewsletter. "My personal opinion is that committees are dominated by suppliers with their own agendas, and confused end-users. By the time ANY standard is announced, it is already outdated by new technology enhancements." "Emerson," he adds, have "cleverly sidestepped the standards process by guaranteeing an upgrade if that's ever needed." A recent analysis of the Wireless Sensor Systems market from Research and Markets suggests that the market for wireless sensor systems will grow rapidly over the next five to 10 years and, depending on the outcome of standardization efforts, could reach between $5bn and $7bn.

With all the excitement generated by Emerson's recent announcements and the anticipation that Honeywell may respond later this month with its own alternative offering, one might have expected Yokogawa to take the opportunity to stick its own oar - or instrument - in. That, however, would be to reckon without the Yokogawa culture. According to Shuzo Kaihori there will be no release of products until they have convinced themselves that their solutions are totally reliable although they expect to have devices on trial with customers early next year. Interestingly, he does not subscribe to the view that wireless may be suitable for monitoring but not for control. Mesh technology

Rather, he argues that monitoring, particularly of critical parameters, requires as high a level of performance as control and therefore that the technology cannot be deployed until it is sufficiently reliable for both. The fact that they're taking broadly the same approach as Emerson with 802.15.4 self healing mesh technology does not mean that there are likely to be any firm announcements from Yokogawa ahead of release of at least draft standards from HART and ISA. That said, however, while the wireless display in the Technology Fair was broadly the same as that presented at last year's 90th birthday celebrations in Tokyo, the low cost sensors aimed at condition monitoring applications were joined by a conventional but fully wireless enabled process transmitter suggesting that, as Kaihori San intimated, bringing product to market is "not a technology issue" and that there is the possibility of products becoming available during 2007.

In the meantime what of the assertion by Yokogawa president and CEO Isao Uchida earlier this year that "the future of plant communications belongs to fiber, not wireless"? No denial that he said it, nor any suggestion that he is alone in thinking that the very high speed, high bandwidth fiber optic technology Yokogawa has developed for the communications industry might be transferable to the process automation arena. Perhaps that's why one got the distinct impression that Yokogawa is not prepared to bet the farm on wireless but sees it as just one more addition to a range of plant communication options. Best indication of that perhaps came in a slide at the end of John van der Geer's presentation entitled 'Future Architecture' which showed Foundation fieldbus, HART and Profibus DP field networks coexisting not just with a self healing wireless mesh network but with what was simply labeled 'optical sensors.'

(IAI - November 2006)

Emerson Process Management duly brought its in-plant wireless technology to Europe, and indeed to the rest of the world outside North America, in the second week of January when it summoned the European press to Bologna in Northern Italy for the launch of the 2.4GHz version of the "Smart Wireless" solutions it unveiled in the US last October (INSIDER, October 2006, page 1).

Why Bologna? It's the birthplace of Guglielmo Marconi and, to put everyone in a suitably wireless mood, proceedings commenced on the Wednesday evening with a tour of the Villa Griffone, the former Marconi family seat where young Guglielmo conducted his experiments which is now a wireless museum. Bologna is also widely regarded as the food capital of Northern Italy and, while some may have found two hours of early wireless technology, much of it in Italian, somewhat indigestible, few would have had problems with the 10 or so courses of Italian country cookery that followed.

Big guns
Perhaps aware that they may have put some European noses out of joint by launching Smart Wireless in North America - and actively trying to prevent the European press from reporting it - Emerson wheeled out a veritable arsenal of North American and European big guns for the occasion: EMEA president Jim Nyquist, Rosemount vice president of technology Bob Karschnia, Rosemount vice president and general manager for pressure Mark Schumacher and EMEA vice president of sales and marketing for systems Steve Brown, all serving to emphasize Nyquist's assertion that this was "the most significant event in this industry in Europe for decades."

In truth, however, other than the fact that it uses the 2.4GHz waveband rather than 900MHz - a difference imposed by the fact that Europe uses 900MHz for, among other things, mobile phones - there wasn't a whole lot in the main Bologna presentations that hadn't been seen in Nashville or, indeed, in the sneak previews at Emerson Exchange in Orlando 12 months earlier. What there was, however, was a stout rebuttal of the key criticisms of its offering made at Honeywell's European user group meeting in Seville in November (see left).

Followers of the ongoing wireless debate will be aware that, while Honeywell has sought to place wireless in a wider context of plant wide communications, Emerson has focused primarily on eliminating the copper between control systems and field devices or, as Nyquist puts it, "unplugging PlantWeb". Both companies have adopted the concept of self healing mesh networks but are seeking to deploy them at different levels, with the Honeywell solution based on a network of mains-powered 'iNodes' while Emerson's is formed by the battery-powered field devices themselves. Honeywell claims that, in the Emerson world, communication loads on the various devices in the network can vary widely, resulting in potentially and unpredictably short battery life.

Bolognese sausage
That, says Emerson, is quite simply a load of old boloney because it ignores the fact that the Dust Networks Time Synchronized Mesh Protocol (TSMP) technology - and Dust Networks president and CEO Joy Weiss was there to confirm it - actually monitors the loading of the individual nodes and 'grooms' the network to avoid just that problem. Even in an extreme case, it would rapidly become apparent if a significant proportion of the network traffic was being passed through a single node and the condition could be alleviated by introducing an additional node. In practice however, according to Karschnia, based on three years' field trials experience, it just doesn't happen in the real world. And, as he told INSIDER, the whole battery issue is greatly exaggerated anyway. Battery technology is advancing so fast that it won't be an issue by the time that the current generation of batteries comes to replaced - and that's without the energy scavenging devices which Emerson already has on trial.

No battery life guarantee
Given that Emerson is saying that battery life will be between five and 15 years, dependent on application, are they prepared to guarantee a minimum battery life of, say, five years? Well, no, says Schumacher. "Because of the variability of the operating conditions, we can't guarantee battery life," he says, although he does say that it can be predicted in a given situation. Tasos Anastasiou, Wireless Technologies leader with BP's Refinery of the Future team at Sunbury, UK, was in Bologna to give the user perspective, based on BP's trials in the US and Euorpe. Interestingly, he says the key criterion is not so much a particular length of battery life but that batteries last between plant turnarounds, however long that might be.

Anastasiou also had his own take on the other major stick with which competitors have been beating Emerson, that of whether its technology can be used for control. Both Honeywell and Yokogawa, for example, have said that if the technology is to be good enough for monitoring in a process environment, it needs to be good enough for control too. Despite suggestions to the contrary, Emerson seems to be saying the same thing - and it points out that its solution addresses classes 1 to 5 of the SP100 use cases which include closed loop regulatory and supervisory control as well as open loop control. "There's no reason why it can't be used for control," says Schumacher. "Customers will want to use it for closed loop control as the technology matures."

That's not quite how Anastasiou sees it, however. While he confirms every one of Emerson's claims for the technology - it really is 99.9% reliable, for example, and it is "very easy to integrate" - he believes that there is some way to go before it can be used in control or safety applications.

Standards
Meanwhile, what of the on-going standards issue? Schumacher certainly seems confident that there will be a draft Wireless HART standard by March and he should know - he's president of the HART Foundation board of directors. Moreover, he says, the SP100 Task Group recommended in December that 802.15.4 be adopted as the default physical layer and that it should concentrate on the field sensor network in the short term.

Target date for a full SP100 standard is still June 2008 and, with technology now in the marketplace, that's more rather than less likely. In the meantime, Emerson has repeated its guarantee to provide an upgrade path to the future standard, although Schumacher scotched any suggestion that the upgrade might be free. More likely it will be a 50% board exchange scheme similar to that offered for HART upgrades in the past. So where does the advantage now lie? Despite a brave showing from Honeywell in Seville, it must still be with Emerson. They've got a network solution in place and products in the field and, despite a chorus of competitive sniping, they seem to be able to convince major users that it's a viable solution. How many of those starter kits they've actually sold since October they're not letting on but the ball is firmly back in the Honeywell court where it's likely to remain until the full release of their network solution, scheduled, so rumour has it, for April.

(IAI - January 2007)

Emerson Process Management claims victory in the race to introduce "In-Plant Wireless for Mainstream Use"

It's just 12 months since Emerson's much vaunted - and, at least by some competitors, much feared - marketing and PR machine found itself wrong footed by its own people when a presentation at the 2005 Emerson Exchange user group conference in Orlando prematurely took the wraps off its wireless technologies. It was that session, for which delegates queued in the corridor and from which Emerson's own people had to be ejected to make room for potential customers, that in effect fired the starting gun in the race to bring real world process industry wireless solutions to market.

Twelve months on, as the Exchange reconvened in Nashville early last week, Emerson was claiming victory as it announced "First In-Plant Wireless for Mainstream Use." And this time there weren't going to be any PR cock ups as the well oiled marketing machine swung into action.

Or so we hear, for the announcement has been made only in the US and, when we requested access to the material handed out to the members of the press present at the launch, we received a polite but firm "No". Ostensibly, the explanation for this bout of transatlantic discrimination is that the launch is to be repeated in Europe in January when, says UK and European PR consultant Charles Lewis, no doubt trying to make us feel better about it, "There will be even more information." Mild subterfuge

In the meantime we've had to resort to some mild subterfuge to obtain the material and, if you're reading this east of longitude 30° W, you'd better stop now if you want to stay on your local Emerson rep's Christmas card list. That of course doesn't mean you can't find out about it - the key press releases are already there for all to see on the Emerson web site. So what was all the fuss about? Well, in technology terms not a lot more than what was so dramatically revealed last year. As the presentations in Nashville confirmed, Emerson is basing what it has dubbed "Smart Wireless" on IEEE 802.125.4 selfhealing mesh network technology. This, it argues, is the key to using wireless in the "steel canyons" of a traditional process plant, providing multiple paths through which signals can find their way from source to destination. Arguably the most satisfying aspect of the technology is that its reliability improves as the number of nodes increases so that Emerson can claim that over three years of site trials with selected end users, principally in the refining and petrochemical industry, it has achieved a proven reliability of delivery of data of better than 99.9%. Moreover, it's claiming that its technology can be deployed without the need for an initial site survey - you pretty much put it in and it works, so they claim.

Reassuringly familiar
From the viewpoint of the field devices, Smart Wireless is something of a nonevent since they talk to the wireless link in HART and so shouldn't know the difference. As a result, Emerson's new wireless devices look almost identical to conventional HART devices. Indeed, any potential user reservations about introducing the new technology are said to be substantially reduced, if not completely allayed, by the fact that the devices not only look but actually are almost identical to the ones with which you and they are already familiar.

If all this sounds too good to be true, that's because there is still a very substantial fly in the ointment. Emerson makes much of its total commitment to standards, which presents problems when there isn't yet a standard to which to commit. This time last year the word was that there would be an agreement on Wireless HART by early 2006. That date has now slipped to February 2007 while the ISA SP 100 committee isn't expected to deliver before summer 2008. And of course it's hardly a secret that all is not sweetness and light within either of the committees, with the suggestion of a division into two camps centred respectively on Emerson and the other pacemaker in the wireless world, Honeywell.

Guaranteed upgrade path
Emerson has clearly decided that, whatever the eventual outcome of those deliberations, neither it nor its users are prepared to wait any longer and is backing the latest introductions with an undertaking, signed by both Emerson Process Management president John Berra and Rosemount Measurement president Steve Sonnenberg, that "Emerson will protect your wireless investment by providing a guaranteed upgrade path to the future industry standard."

How that guarantee will be fulfilled if the eventual standard isn't based on a self-healing mesh network is not immediately clear but would presumably pose a huge problem to Emerson.

Nothing daunted, however, its initial in-plant Smart Wireless solutions include pressure and temperature transmitters, with other parameters and devices said to be following shortly; 900Mhz wireless gateway, with 2.4Ghz gateway for Europe, Asia and Latin America following in January; OPC, Ethernet and Modbus integration with host systems, with native wireless interfaces for DeltaV and Ovation also following shortly; and a direct wireless interface for the AMS Device Manager asset management application.

Also coming shortly is the means to integrate what is called the "Stranded Diagnostics" of the 20 million or so HART devices currently installed in plants which don't support a digital architecture.

If you're using some form of point to point or star topology, you can of course dip your toe into the wireless water with a single device. However, you can't have a mesh of one so Emerson is hoping to persuade wireless virgins to take the plunge by shelling out a minimum of $15,000 on a 'Wireless SmartPack Starter Kit' comprising the 1420 Wireless Gateway, a 25-tag license for AMS Device Manager and anything from five to 100 Rosemount temperature, pressure, level or flow devices. The deal also includes 'SmartStart' installation services under which Emerson's own experts help with first startup, providing a full network health assessment to ensure robust communications and verifying device functionality through the chosen output. "Historically, customers must commit significant resources for engineering, design, configuration and startup when trying new technologies," said John Berra. ". . . Wireless SmartPack gives mainstream customers an out-of-the-box network which they can apply anywhere, anytime with very little engineering time required. It's easy, and it works straight out of the box."

The elephant in the room
Even at a few thousand miles distance, it's not difficult to discern that the elephant in the room at the Smart Wireless launch had Honeywell written all over it. Following the latter's presentation of its wireless strategy at its own user group meeting in Phoenix in June and its favourable review by ARC (INSIDER July 2006, page 9 and September 2006, page 1 - Subscribe to IAI ansd ask for back copies!!), Emerson were keen to stress that this wasn't 'vaporware' and that you could place your orders today - even if the anticipated demand might mean you had to wait a little for delivery.

It was also stressed in the presentation that Emerson had conducted extensive lab and customer trials of both Cluster Tree CSMA (Carrier Sense Multiple Access) and Mesh TSMP (Time Synchronized Mesh Protocol) technologies before finally opting for the latter. Equally the suggestion that Honeywell might have the edge because of its longer experience with wireless was countered by Emerson's own claim that it is already a wireless leader with, according to ARC, a 19.9% share of the wireless market.

On the key issue of whether wireless is ready for control Emerson, in contrast to Honeywell, still believes it is too early to commit but claims that its solution addresses all five of the SP100 use Classes, including those for closed loop control, although only classes 3 to 5 have been proven in use by customers.

More to come This autumn will inevitably see a flurry of further activity, not to mention the inevitable claims and counter claims, on the wireless front. Next month Honeywell has invited the European press to its European User Group meeting in Seville where it's difficult to believe it will miss the opportunity to respond to the latest Emerson introductions while Yokogawa has an event in Prague later this month which should provide the chance to find out whether president and CEO Uchida San still stands by his statement made earlier this year that "the future of plant communications belongs to fiber, not wireless." And that's all before we Europeans are let in on the Emerson secrets in January. Remember, you didn't read it here first!

• What galvanized delegates' attention at last year's Exchange was the contribution from "a major refiner" who revealed that extended trials of Emerson's technology hadn't just proved that it worked and worked reliably but had demonstrated reductions in installation costs of typically 90%. Put another way, that represented an order of magnitude reduction in the cost of installating a $1000 transmitter from $10, 000 to just $1000. Presumably it's now safe to reveal that the refiner in question was BP since David Lafferty, senior technology consultant in Digital & Communications Technology with BP's Chief Technology Office was the star turn at last week's launch. Lafferty confirmed the experience reported last year when he said that "We've found that this wireless technology enabled us to do things we simply could not do before, either because of cost or physical wiring obstacles . . . Emerson's wireless approach is flexible, easy to use, reliable, and makes a step change reduction in installed costs."