Real promise for OLE

Development of the OPC open standard for process control is gaining pace with the emergence of the first genuine commercial software releases. George Paloczi-Horvath reports

Between now and early summer, significant developments are promised for OPC, the emerging software standard that allows non-proprietary business software to communicate with specialist process industry software.

Developments involve revisions to the initial OPC Release 1.0 specification, issued last November, which will lead to better practical use of the OPC standard.

OPC’s history is complex: software companies serving process industries decided to extend Microsoft’s Object Linking and Embedding (OLE) manufacturing software to offer a new `OLE for Process Control’ (OPC) standard. The development was led by the OPC Task Force, which is a group of companies within a larger OPC Foundation which combined to develop the new software interface.

Basically, OPC allows software firms to avoid the tedium of writing complex custom software interfaces to permit data exchange between ordinary business software on one side and specialist process control software (and the machines it controls) on the other.

The OPC Foundation’s aim is `to develop an open and interoperable interface standard’. This would be based on the functional requirements of OLE/COM (Common Object Module) and DCOM (Distributed COM) technology.

Initial OPC task force meetings took place two years ago and a first draft standard was published in December 1995. Last August the final OPC requirement was issued, which led to November’s release of the OPC Release 1.0 specification.

Demonstrated at Chicago trade fairs last October and in April this year, this spec served the purpose of familiarising industry with OPC but the 1.01 and 2.0 specs will provide more user-friendly standards.

An executive involved in OPC’s development explains: `There are many companies saying they will have OPC-compliant products, but do not yet really have them except as demonstrator products. Early users were companies’ technology development departments. The first release was targeted at these departments. We’re now at the stage when we’re seeing the early deployment of OPC-compliant technologies.’

The significantly more advanced Release 2.0 is promised for early summer, but could emerge in two separate releases.

Adrian Wise, managing director of USData Europe, an OPC Foundation member, says Release 1.01 is a more complete evolution of OPC 1.0, encompassing some additional, improved functionality. One of the most important goals in the 1.01 spec will include definitions of how you interface low-level data access. This contains USData containerisation technology, which embodies the ability to put one software object within a larger software object.’ In this context an `object’ is a module of software with defined characteristics which dictate what it has to do and how to pass information to and from it.

`With this release, we know we’ll have a standard that will allow us also to offer products,’ Wise says. `OPC extends OLE into the realtime world, allowing you to control events. It could be an interrupt you need to respond to, or it could be a human or timer event.’

Another senior OPC-linked executive put it thus: `In process control manufacture – in the food and chemical industries for instance – the traditional way of controlling that environment is through distributed control software.

`OPC’s advantages are obvious: you don’t have training issues, with the advantages of using an open, well-understood operating system. The whole process control industry is moving to open systems.’

Process industries now use fieldbus electronic links to control valves and instruments and other hardware. Previously this hardware was driven by a mainframe computer but is now increasingly controlled via fieldbus through localised PCs. The problem was that special software had to be written to allow those PCs to link with process hardware.

With the increasing introduction of PCs, an innovation such as OPC became essential to take the process control industry beyond fieldbus. `Suddenly, you’re buying low-cost, well-understood software products and using them for your production problems,’ the OPC specialist says.

`You can then link the data that’s coming out of the field and put it into software you understand. For the first time, you have realtime process data. A process plant generates a huge amount of data; managing that data is the challenge.’ But he warns: `Nobody has cracked some problems properly. The industry has struggled for 10 years to get fieldbus sorted. The OPC element is just one part of the whole story.’

Release 2.0 of OPC will be released this summer. Alpin Chisholm, chief technical officer of Intellution, a firm in the OPC Task Force, says this release will include support for Microsoft Visual Basic 5.0; that is, the language used in PC networks to develop `scripts’ which allow the automatic execution of multiple, repetitive tasks.

David Rehbein, the OPC Foundation president, is from Fisher-Rosemount, a major OPC Task Force company already offering OPC-compliant solutions such as Delta V. Rehbein says that some parts of Release 2.0 will include `purely cosmetic changes’ and others the `automation stuff’ involving Visual Basic 5.0.

But USData’s Wise warns that OLE is misunderstood. `People shouldn’t run off and think it’s a panacea. While it’s based on a Microsoft specification, the company’s position on OLE isn’t clear and it might come up with something new. We like to think that OPC is going to be an industry standard that isn’t going to change, but we all know that Microsoft has the ability to change industry standards: it leads and the rest of us have to follow. The other caveat is that there are question marks about OPC performance. Using industry standards may be a good way to produce interoperability, but it’s not a quick way.’

But a senior OPC-linked executive says markets make standards, not the other way round. `What Microsoft is doing with OLE is driving a standard into the marketplace and coming to terms with a recognition that the future of process control lies with the PC.’