Issue link: https://iconnect007.uberflip.com/i/1494346
30 DESIGN007 MAGAZINE I MARCH 2023 associated with machines and soware solutions. Automation technology is developed using that know-how, so that key manufacturing dependencies on specific people are avoided. e value from that experience is easily realized in their choices of machines and soware tools. What I'm hearing is that tribal knowledge is valuable, but it will fade away unless we capture it. ere is a simple, but unfortunate pattern that emerges when we try to transfer knowledge from one person to another. Initially, you train someone to perform a certain role, and over time, they become absolute experts in what they do. e time comes, however, for them to move on. Everyone likes the opportunity for career progression. eir replacement must be trained. Will 100% of their knowledge and experience be transferred? No. Some of that knowledge may not even be valid today, and that's where it gets tricky. e priority is always to transfer the part of the knowledge that is in use, relevant to the current operation, which represents a fraction of their total knowledge. e same cycle then repeats, reducing the amount of knowledge and know- how overall. is is a natural process, as there is limited time in which to do those transitions. e worst part is that the operation appears to be operating as required but is actually becoming totally inflexible. ose performing their tasks know less and less about why they're doing it. ey are only doing what they've been told and are not confident to deviate from that operation in any way because they don't know the reasons behind the tasks. Inf lexib ilit y breed s inef f ic ienc y, becaus e where there needs to be change and evolution in the way that things are done, the ability to adapt and optimize are simply no longer there. Who's responsible for deciding which knowledge is transferred? It follows the law of diminish- ing returns. If the manager wants to make sure that the training is more comprehen- sive, whether for people mov- ing internally or being replaced from the outside, resistance to do so increases as the relevance to the immedi- ate operation reduces, making it a lower prior- ity over time vs. daily operational challenges. Clearly, the responsibility for drawing the line falls on managers. It's difficult to balance these needs for training against the needs of continu- ous operation. You can't blame them for that; it is, I'm afraid, a natural process. Obviously, there is value if the tribal knowl- edge is captured correctly and validated within software tools. You still have to validate things one way or the other. I'm thinking this applies even down to the layout of a board, because a PCB can be designed an unlimited number of ways. Exactly. I remember my relatively few days of designing printed circuit boards years ago. It was hard work having to engineer every aspect of the layout yourself, with only simplistic word-of-mouth design rules. Over the years, we've seen the know-how of design layout becoming increasingly automated within so- ware, as know-how is incorporated into the soware little by little. Over time, it became easier and quicker to perform the PCB layout. It soon got to the point where there was no longer any need to do all that initial low-level layout work. Today, you simply rely on the soware tool's algorithms to put the compo- nents where they ought to be, and then route the layout, basically, replacing all the boring, Michael Ford