By Eugene le Roux, mechanical engineer, one of his regular online articles on general project management principles.    

Would you like to join the discussion on some aspects of configuration management? 

Eugene le Roux, retired mechanical engineer

Eugene le Roux, retired mechanical engineer.
© RACA Journal

If so, let’s consider the following:

What could be done to prevent unapproved changes from being implemented unofficially, and what is the impact of such practices?

Given that hardware changes inevitably lead to a domino effect of repackaging of the surrounding components, how must these changes be scheduled to have the optimal benefit for the product? Should IOWs (Integrity Operating Windows – sets of limits used to determine the different variables that could affect the integrity and reliability of a process unit) first be accumulated until a meaningful amount can be executed in a single repackaging and refit programme, or is piecewise implementation acceptable?

Sometimes changes are urgent because of a scarcity of supply to the production line, or because of an emerging safety concern. Normally a change must be considered against its impact on cost, system reliability, impact on all the other project documentation, including drawings, logistic support, and more. This all takes time which one often does not have. Does your configuration plan allow you to implement an urgent change by means of a deviation, and the follow it up by an Engineering Change Request (ECR), in which everything could later be evaluated thoroughly?

But what about the risk involved in implementing a fully unresearched change? Who must approve this risk, and does the configuration plan require the reasoning behind the decision to be motivated, and documented for later reference?

May I ask you to add to this discussion so that we could all learn different ways of handling changes?