Trust Engines to Preserve the Quality of an Operational Decision System

After a long period where automation was the most important focus, the place of the human operators has been revalorized by the modern production system in order to reach higher production agility. In this context, we have proposed an assistance system that provides to the operators of a unit a limited subset of the most productive tasks, the new optimization being computed according to the operators' final decisions. However, this kind of system may be flawed because it relies on the assumption of perfect human collaboration. In this paper, we present the trust issues that are created due to such human involvement in the operational decisions. To address the impact of these issues, we integrate a computational trust engine into our decision system in order to determine the trust level of an action carried out by an operator.