In contrast to {p-O}R where the score can be high when one element is rather low if the other is exceptionally good, there are situations where all elements must be present. This necessitates using the {p-AN}D operator.
Fuzzy logic depends on membership functions that may be rather arbitrary judgements. Nevertheless, most real-world problems are not yes-or-no or black-or-white, and a systematic approach to working with fuzzy information opens new vistas for analysis and control.
With both ordinary expert systems and fuzzy control systems, rule activity charts provide valuable insights. These charts show when a rule fires and whether it was weak or intense. Although the designers of the controller may think that certain control features are more important than others, the rule activity chart tells the true story. In the case of fuzzy control of the activated sludge process, it was anticipated that SWR would dominate but actually it was RRSP that frequently changed. The implications are that the rules that "fire" most often should get careful examination. Furthermore, there may be need for a threshold for rule activation so that it does not fire unnecessarily. One concept is to deactivate a rule when there is a poor match between the current fuzzy measurement and the input condition. The fuzzy qualifiers may be "weakly satisfied" or "strongly satisfied".