Rough sets and maintenance in a production line

The great volume of information used/obtained in industry nowadays produces the need for systems that are helpful in decision-making. An area of special interest, because of its influence on productive capacity, is maintenance, where the great quantity of influential variations and available facts (including real time) make the analysis almost impossible. Using the concept of rough sets, we present a comprehensive (but real) application where the information system is reduced so as to get a minimum subset of attributes without loss of quality; a set of nine decision rules is obtained.