Modelling of the decision-making process of jobs priorities definition under conditions of limited resources: case of a service company

This article focuses on the problem of decision-making regarding the prioritization, selection and scheduling of maintenance and repair jobs on complex industrial equipment under conditions of limited resources for the case of a service company. The author reviews modern approaches to predictive maintenance and jobs scheduling, namely job shop scheduling problem (JSSP), noting a gap in combining these two tasks for the case of service companies and absence of its strict scientifical substantiation. As the first step, a mathematical model is proposed which formulates the problem as an integer linear programming problem (ILP) with the objective of maximizing revenue and minimizing the total duration of the work execution under the limited resources of the service company. The scheduling part in based on the flexible JSSP, one the latest developments of JSSP, which allows to consider business-processes of the services company, which operates a fleet of almost identical mobile workshops. The peculiarity of the researched model is usage of time- indexed variables. Since the revenue is unknown at the moment of decision making, it is modelled with type-2 fuzzy logic approach, which mimics the real expert decision making. Application of type-2 fuzzy logic, in contrast to the type-1 fuzzy logic, allows to consider both uncertainty of experts in preliminary revenue evaluation, which could be modelled by type-1 fuzzy logic, and vagueness in determination of membership function. Directions for further research are outlined, among them numerical experiments on real and synthetically generated data, and further multi-criteria optimization considering equipment downtime. In further research it will be necessary to overcome the increased computing complexity of type-2 fuzzy logic. A few ideas are proposed to avoid this obstacle.