There is nothing new except what has been forgotten.
Marie Antoinette

Independent of how much data you collect and independent of the form you use to collect it there will always be a question that cannot be answered. Take it for granted that five, ten, or twenty years in the future you will be saying, "If I had only collected that one piece of information;" or "If only I had collected the information in a format that could have been more easily sorted". This is the result of finding new elements that are part of patterns not currently recognized for their importance.

An example of such an element is contaminants in oil. The contaminants have always been there and we have had the tools needed to measure them for quite some time. Despite that facility, it is only recently that many companies have begun to apply sophisticated oil analysis techniques to condemn oil. Most oil is still changed based on time or mileage and not the condition of the oil.

It is impossible to avoid the regret associated with not having collected all the data, but it is possible to minimize the regret by thinking of all the ways you may use the information, and what data will answer reasonable questions now and in the future. Here are a few of the questions you may be asked:

1. Is this a weather-related failure? If so, am I tracking weather conditions at the time of failure?

2. Is this kind of failure limited...