مدیریت عدم اطمینان در محیط های تولید برنامه ریزی منابع سازمانی (ERP) تحت کنترل در بنگاه های کوچک و متوسط (SME ها)

This research examines how and to what extent uncertainty disturbs the SMEs in the manufacturing sector, which plan and schedule their production using MRP, MRPII or ERP system, and proposes to apply a business model to manage uncertainty. This environment is termed the ERP-controlled manufacturing environments. A comprehensive literature review found that various buffering and dampening techniques have been used to tackle uncertainty. The results of which show consistent late delivery performance, reinforcing that there is a clear shortage of knowledge and guidance on how to tackle uncertainty, particularly in these SMEs. In this study, a business model that enables diagnosis of underlying causes of uncertainty is applied through a questionnaire survey in order to identify the types of underlying causes that are more likely to result in late delivery. The survey results provide the SMEs with a reference on the underlying causes of uncertainty that must be tackled with higher priority. Simulation modelling and experimental study of the underlying causes of uncertainty on late delivery based on a real case study verify and validate this suggestion.

Traditionally, Material Requirements Planning (MRP) and Manufacturing Resource Planning (MRPII) systems are used by large enterprises as a production planning and scheduling tool. Over the last 10 years, a new system has evolved from these systems, namely the Enterprise Resource Planning (ERP) system. These types of systems are now seen as an enterprise wide integrated information system. In an enterprise context, their main aim is to provide information to/from finance, accounting, sales, marketing, planning, production, purchasing, human resource, logistic and distribution so that the entire organisation process from receiving a customer order, to manufacturing, and final delivery is structurally and systematically handled by the system. In a production context, their main aim is to produce a reliable Planned Order Release (POR) schedule in order to meet the delivery due date. Whatever system is chosen, it must be capable of performing within an uncertain environment. Uncertainty, in this context, is defined as any unpredictable event that disturbs the production process in a manufacturing system that is planned by MRP, MRPII or ERP system (Koh and Saad, 2003a). Some researchers referred to uncertainty as a form of disturbance (Lindau and Lumsden, 1995; Frizelle et al., 1998; Saad and Gindy, 1998). Regardless of the terms that it is being referred to, overall, these researches examined a variety of buffering or dampening techniques to minimise the effect of uncertainty.

A business model for diagnosing the underlying causes of uncertainty in manufacturing enterprises that use MRP, MRPII or ERP for production planning and scheduling was applied to study the underlying causes of uncertainty in ERP-controlled manufacturing environments in SMEs. A questionnaire survey was carried out to identify the causes that are more likely to result in late delivery in those SMEs. ANOVA results showed that poor supplier delivery performance, schedule/work-to-list not controlled, machine capacity shortages, finished product completed—not delivered, unacceptable product quality and engineering design changes during/after production have significant effect on late delivery. The interactions between unacceptable/urgent changes to production schedule and poor supplier delivery performance; and unacceptable product quality and engineering design changes during/after production yielded additional level of late delivery. To test whether these results could be used as a reference for the SME manufacturers in tackling the significant causes of uncertainty, simulation modelling and experimental study using a real case study were carried out. The significant causes of uncertainty and the significant interactions between uncertainties found in the survey were modelled. A half-factorial design of experiments was performed and 485 replications of the experiments were run. The data used in the simulation experiments was derived from a medium-sized transformer manufacturer. The size of the parameters in the simulation model could be generally characterised by: two-year MPS data, 434 different parts and 50,000 orders in the POR schedule.