Object-Oriented Bayesian networks for complex quality management problems

Quality management and customer satisfaction evaluation can be difficult tasks to perform when processes involve multiple production lines or provide multichannel services. As a consequence, the top management needs to check the quality from different perspectives and to evaluate the improvement strategies at several levels. To this aim, we propose to use Object-Oriented Bayesian Networks by which different quality aspects and evaluations can be integrated in a unique framework allowing to analyse improvement strategies in real time. We show, by an application to an internal-customer satisfaction survey, how to combine several areas of satisfaction and how to evaluate the impact on the global quality of improvement actions developed in one or more areas.