In the chemical industry, faster doesn't always mean better

Real-time control has turned up in controllers, analyzers, control systems and more. In the end, real-time information is only as valuable as an organization makes it.

By Nick Basta, editor at large

Share Print Related RSS

I’ve always been a fan of automation, in nearly all its forms. Automated systems are generally faster, more accurate and less expensive than alternatives. In process control, the point is a trivial one. Automation is the key driver to running more complex manufacturing processes, and to running them closer to theoretical thermodynamic limits.

Along the way to our current highly automated manufacturing scene, an interesting metaphor has evolved: real-time control or automation. There are real-time controllers, real-time analyzers and real-time control systems. Now, the gurus of the information-technology and computing fields speak glowingly about real-time scheduling of business activities, real-time inventory control and real-time business management.

As these phrases keep popping up, one wants to say, “Wait a minute. Let’s get real about this.”

As one looks into the matter, it becomes clear that there are at least two kinds of “real time.” One is defined — pretty rigorously at that — in the realms of computer science and microelectronics. The other is so broadly used and misused in common speech that it is becoming nearly meaningless; at best, it means fast.

In the computing world, real time indicates, among other things, that an application can perform instructions at a rate at least as fast as that of incoming data. (Interestingly, this sometimes can be quite slow, even taking a matter of seconds.) It also means that the system is deterministic — that certain things will happen at certain times regardless of what else is going on. To quote from a statement from the Real Time Java Group, “Real-time systems require strong deterministic guarantees and/or control in the areas of thread scheduling, synchronization overhead, lock queuing order, class initialization, maximum interrupt response latency …” (This is very impressive and makes it even more mysterious as to why I’ve yet to have a PC that can keep time accurately!)

In the world beyond microelectronics, “real time” implies that something can be communicated or reacted to now, and not a moment later. However, even if information is delivered very quickly, responses rarely occur instantaneously. You can watch real-time weather on the Internet, but you’re better off looking out a window to decide whether to take an umbrella.

In process control — and the management of chemical businesses — the meaning of real times falls between these two extremes. According to Peter Martin, an automation guru at Invensys Corp., Foxboro, Mass., real time used to be defined simply (and somewhat arbitrarily) as one-quarter of the time cycle of a control loop. That gave a controller enough time to react to a signal and send an instruction through the loop before the next cycle. Today’s microelectronics make reaction and communication times much faster, but a loop still is constrained by the physics of the process it is controlling. It doesn’t help to sample a tank level 5,000 times a second if it takes a day to fill the vessel.

Automation vendor OSISoft, San Leandro, Calif., for one, during the past couple years has been making a big deal of what it calls “real-time performance management” (RTPM). Its systems enable a key performance parameter, such as the electrical consumption of a production line, to be extracted from a welter of process data and delivered to another system, such as one that recommends changes to operations based on the variable pricing offered by some power utilities. Plant managers can then make meaningful adjustments to save power costs.

To gain the benefits of something like RTPM, however, depends upon the willingness of an organization to understand its systems well enough, and to delegate responsibilities low enough, so that someone can take advantage of the information. Unfortunately, a lot of the potential of real-time business management is lost because companies won’t adjust their business processes. 

In the final analysis, real-time information is only as valuable as an organization makes it. Smartly managed companies will win over smartly automated companies. That’s about as real as real can get.

You can contact Nick Basta at Nbasta@putman.net

Share Print Reprints Permissions

What are your comments?

Join the discussion today. Login Here.

Comments

No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments