It's now almost four years since what was then called Invensys Process Systems (IPS) launched in InFusion, what it claimed was the world's first enterprise control system or ECS at a spectacular event in Boston's John F Kennedy Memorial Library. Much has changed at what is now called Invensys Operations Management (IOM) in the interim, not least, as well as its name, its CEO―twice―and its relationship with Wonderware. Intriguingly, INSIDER's May 2006 issue had IPS' then-CEO Mike Calliel explaining the need for the continued separation of IPS and Wonderware because of the potential danger to the latter of too close an association with the former. Clearly his successor but one takes a rather different view.
Changed Perceptions
But it's not just names and the organization that have changed in four years and most dramatically in the past year or so. Perceptions too have altered significantly, both of the role an enterprise control system should play in the space between the real-time world of process control and the transactional world of the enterprise and, more specifically, of how InFusion, both as a deliverable and as a brand, should be positioned in Invensys Operation Management (IOM)'s portfolio of offerings.
As Peter Martin, now IOM's vice president of business value generation, explained when he and director of commercialization, Steve Garbrecht, briefed INSIDER on the latest release of InFusion late last month, four years ago the emphasis was very much on bringing information together from disparate sources across the enterprise, many of which might and indeed all of which could be non-Invensys systems, and passing the consolidated information up to the enterprise. Indeed that was underlined by the significant role SAP played in the 2004 launch and by the presence at it of former Lighthammer CTO and then SAP vp of manufacturing applications, Rick Bullotta. (Bullotta, incidentally, was subsequently briefly CTO at Wonderware before moving on to found Burning Sky Software, a pioneer of collaborative, ‘real-world aware' applications, where he is chief strategy officer.)
In Martin's view "Something has changed significantly in the last four years." Whereas the assumption had been that plant-level systems would report to the enterprise level where things were comparatively stable, that no longer stands up. "The whole business world is out of control," says Martin somewhat dramatically, citing the example of energy prices which a few years ago could be negotiated on a monthly or even annual basis, but which now can change by the day or even the hour. "Business," he argues, "has become a real-time control problem," the implication being that it needs real-time control systems and control engineers to run it. Instead of there being a pull from ERP for reporting—"Reports don't actually do anything," observes Martin—the pull is in the other direction, to bring decision-making down into the real-time world with the ultimate objective of creating what IOM president and CEO Sudipta Bhattacharya, among others, has called "a real-time enterprise."
I/A integration
As we have already noted, however, it's not just the perceived role of the ECS that has changed. Go back four years again and the emphasis was very much on InFusion's openness and, hence, on its ability to integrate with any lower-level system from any vendor, of which IOM's I/A Series DCS was just one. But if it was equally easy to integrate with any system, then you could also say it was also equally difficult, and one of the pressures has been to make that integration much simpler and, in effect, pre-engineered.
If you're looking to do that, then it's pretty obvious that you're going to start with your own systems and, while that doesn't mean that it's more difficult than it was to integrate with third parties, it's difficult to avoid giving the impression of a change of emphasis. And, in fact, that change goes further because with the new release of InFusion also comes two versions or subsets, InFusion Control Edition and InFusion Foundation Edition, one of which is entirely specific to I/A, while the other, though more generic, provides IOM-specific capabilities, notably to integrate its SCADA offerings for the oil and gas, pipeline and utilities sectors.
Higher level capabilities
Control Edition essentially extends the ArchestrA technology within InFusion to I/A, adding a broader set of higher level capabilities than IOM claims are normally found in a conventional DCS. Specific new functionality includes easier and faster configuration of control in the field with Foundation fieldbus―extending InFusion to the field, as Garbrecht puts it―and access to InFusion configuration data through SQL Views. The combination of Control Edition with the latest release of I/A also supports new levels of cyber security in compliance with the full gamut of current security standards including the NERC CIP requirements.
What this all actually represents, says Martin, is the latest stage in an evolution which can be traced back all the way to the original bringing together of the constituent elements of IOM―Foxboro, Triconex, SimSci-Esscor, Wonderware and Eurotherm―under the auspices of Invensys' predecessor, Siebe. In the Martin version of the events of the past decade or more, and in contrast to the other automation vendors which have grown through acquisition and which sought to integrate their organizations before integrating their technologies, IOM and its predecessors put how to pull together their technologies as their first priority. That, he argues was the strategy underlying the development of ArchestrA as an "industrial wrapper" around .NET which could provide the infrastructure for that integration.
With ArchestrA providing the basis both for Wonderware System Platform and InFusion, and, hence, for the eventual integration of the entire product portfolio, the formation of a common development organization was a logical next step and an essential precursor to the eventual integration of the former IPS and Wonderware to form IOM last year.
If all that sounds like something of a post hoc reinterpretation of the Invensys story, there's not much doubt of what we're going to see in the next stage. "One of our biggest struggles now," says Garbrecht, "is to bring 50 different product lines into a single product line. We're converging them into a single product line, which is InFusion." The question, he says, is "How quickly can we jump to InFusion as the primary product line for the company?"
Standard templates
Just how extensive that ambition now is can be judged from a PowerPoint slide showing Wonderware, Triconex, Foxboro and Eurotherm products and applications all coming under the heading of the InFusion ECS by virtue of the fact that they are now linked into InFusion through standard object-based templates. In effect, the argument seems to be that any product or application which is InFusion-enabled in this way becomes de facto a part of InFusion. On that basis everything from the Wonderware Intelatrac mobile operator solutions to Triconex safety systems, and from Foxboro I/A to the Eurotherm-based Foxboro A2 becomes part of InFusion ECS 2.0. Whether that argument would be pursued to include third-party products is another matter, though the assertion remains that the ArchestrA Object Toolkit and the InFusion Integrated Engineering Environment make such third-party integration almost as straightforward.
None of that however means a total abandonment of the goodwill associated with the individual brands. IOM is caught in much the same dilemma as its competitors have been in deciding whether, when and where to retain or abandon its legacy brands and that can lead to some confusing anomalies. Hence Eurotherm hardware and software comes into the fold as InFusion PAC while Foxboro I/A comes in under its own name, but under the umbrella of InFusion Control Edition and seemingly everything on the SCADA side appears to become InFusion SCADA.
Native integration
Garbrecht argues that enterprise control systems have always existed, but that the tools available to create them rendered them brittle and almost impossible to update and extend without breaking the whole system. What changed with InFusion was that, with the adoption of open, object-based technologies, integration became potentially a standardized engineering procedure which could be executed repeatedly without the danger of damaging the existing structure. InFusion 2.0 takes that concept a stage further by making the integration of IOM's own products "native," while responding to what Garbrecht claims has been customers' desire for vendors to move away from a purely product focus. In short, "Every aspect of the ECS has been enhanced."