A model for advanced planning systems dedicated to the Engineer-To-Order context

Abstract

This paper presents a new mathematical formulation for planning and scheduling activities in the Engineer To Order (ETO) context. Designed from an Advanced Planning System perspective, the proposed formulation not only schedules production operations but also takes into account the assembly, design, engineering, and validation phases. The definition of resources is thus generic and enables to model employees, finite capacity machines, and consumable materials. The definition of operations allows to represent short production operations, respecting precedence relations representing the assembly of elements, but also non-physical activities. Non-physical activities are longer, subject to validation, and applied once for multiple identical elements. Furthermore, to integrate planning and scheduling, the proposed formulation is not limited to time-based objectives but also considers financial and organizational aspects. The experiments carried out on instances with up to 100 operations show that our model performs well and requires reasonable computing times. Besides, we propose an ETO strategy that tends to validate the design of non-standard and highly uncertain items first and delay their production or purchase. Our integrated model governed by the proposed ETO strategy is compared to a model that mimics decision processes in existing industrial systems. The comparative study and experimental results highlight how this strategy yields robust integrated solutions that offer a good trade-off between the wastes caused by unpredictable changes in the BOM and the projects completion time.