Modelling the operator know-how to control sensory quality in traditional processes

Traditional foods are generally manufactured in small factories where operators often play an important role: (1) to make on-line evaluations of the properties of foods and/or (2) to adjust the process variables to ensure a smooth running of the process and respect of the quality requirements. The paper presents the methodological guideline we have developed to manage the expert-operator knowledge for controlling the sensory quality of food products. It involved several steps: collection of sensory measurements, instrumental measurements and heuristics controlling rules; modelling of the operator know-how by using suitable mathematical tools such as fuzzy logic or expert systems; development of decision support systems, easy to use by the operators. The principles and the results of the method will be illustrated by examples of traditional processes: dry sausage processing and biscuits aeration. As a conclusion, the main interests of the approach are underlined: traceability of the practices, safer measurements and practices, formation of inexperienced operators, increase of the reliability in the decision of the operators and valorisation of their role. (C) 2007 Elsevier Ltd. All rights reserved.