At breakfast this morning, Eddie Habibi, founder of PAS and I were talking about the importance of the type of change management, configuration management and geneology management that his "Integrity" product does. One of the topics of our pretty far ranging discussion was the failure, so far, of the "knowledge base capture" campaigns that companies have run for the past 15 years trying to suck the institutional knowledge out of the "grey headed guys" and get it down into some sort of database that can be used for training and decision support.
The problem, we agreed, was the fact that the grey headed guys don't know what they know. Institutional knowledge is diverse and in some cases intuitive. So you can't just say, "tell me everything you know about digester control." They respond, "well, uh..." They know what to do, but in many cases they cannot articulate the why.
The next thing people have tried is to get them to do electronic post-it notes...when you do something, open a note window and write down what you did and why you did it. So in an emergency the operator is going to remember to do that? Somehow I don't think so.
One of the biggest problems with DCS and SCADA systems is the lack of forensics. This is a problem in cybersecurity as well.
What we need is a way to capture the information as they perform their jobs in the same way they do it-- that is, intuitively, Then we can have them review and comment on it when they aren't trying to keep the plant from blowing up.
Integrity, based on the concept of the automation genome, is headed in that direction.