PC? Perpetual change...

When I was at the CIM show towards the end of last year I got talking to one of the other trade magazine editors. In the past he has made great use of a Psion PDA, even for typing notes during press conferences, but he is now using a pocket PC instead. His opinion was that the expandability of the pocket PC is much better, but that the unit suffers from frequent crashes.

Fair or otherwise, the accusation of instability is one that is often levelled at Windows (CE, 98, NT, or any other flavour). So, when it comes to PC-based control systems, it is hardly surprising that many users and suppliers believe that the alternative Linux operating system has a future.

Paul Herron, product manager for displays and industrial computers at Rockwell Automation, has commented: "Linux is seen as an excellent, low cost alternative to other, more expensive operating systems such as Windows NT. However, Linux is still viewed by some to be not as well supported and therefore more risky. Regardless of these factors, Rockwell Automation expects to see an increase in its use as an operating system for integrated ruggedised/industrial-grade computers."

But we need to be careful with our terminology when it comes to PC-based control. For example, 'PC' used to be shorthand for an IBM Personal Computer and, at that time, we were careful to refer to PC-compatible computers for the machines that, generally speaking, did the same but cost rather less. These days 'PC' has become a generic term that is used to refer to everything from desktop machines to ruggedised/industrialised machines, solid-state devices using the Windows CE operating system, and single-board computers.

Broadly-speaking, therefore, PC-based control now extends from the traditional SCADA, data acquisition and process control applications to much lower-grade tasks, such as human-machine interfaces (HMIs). In some cases PCs are also taking over from PLCs.

A further blurring of the boundaries has occurred following Mitsubishi's launch of the Qn series of PLCs at the back end of last year. Alongside a PLC CPU, users can install a PC CPU on the same backplane, as well as a motion controller CPU. The common backplane saves space in the control cabinet but, more importantly, it also solves all the worries about how to network the devices together.

I won't enter into the long-running (and ongoing) argument about whether Windows CE 3.0 is truly a real-time operating system because plenty has already been written about this topic and, to a large extent, it depends on your own definition of 'real-time' and how much of a delay is acceptable to you for a particular application. But there is no denying that Windows CE is being used more widely in the industrial environment thanks to several other factors: the common user interface and development tools as for desktop operating systems; CE's modular nature that helps to provide a robust solution for industrial applications; ease of integration with PC-based controllers installed elsewhere in the plant; and easy access to the internet.

So where does all this leave PC-based control in industry? The scalable nature of the Windows operating system certainly is an advantage, though many users will only use Microsoft's products grudgingly. There are alternatives, such as Linux, so PC-based control can be achieved without a Windows OS. And despite Windows CE-based devices starting to penetrate the markets traditionally dominated by suppliers of PLCs and HMIs, these classes of product still have their role to play in industrial automation projects. Much has changed in the last few years, and it is certain that there will be more change in the future; last year's optimal solution to a problem may not be this year's, and next year's could be different again. Stay informed; be aware of the ever-evolving state of the art.