Modeling of Decision Making Process for Product Service Failure Diagnosis

Product malfunctions in service are addressed by service centres that diagnose the problem and make decisions on component repair or replacement actions. Feedback in the form of service centre performance in diagnosing and reparing product service failures is an important indicator of percentage of ‘First-Time-Right’ repairs, which can be improved by analytical modeling of decision making process related to fault diagnosis and repairs done by best service centres. This research proposes to develop a analytical model of service centres’ decision making process as IF-THEN decision rules which link repair actions with product pedigree. The proposed analytical model is based on Rough Sets theory and identifies decision rules which give statistically significant association between product pedigree and repair actions. Analysing the decision making process of service centres is important in developing pre-alerting fault diagnostic rules, recommending best practice diagnoses and sending feedback to OEM for potential changes in design and manufacturing. A case study from automotive warranty demonstrates the methodology.