The effect of human learning and forgetting on fuzzy EOQ model with backorders / Nima Kazemi

Inventory planning is a repetitive task, often characterized with inventory planners experiencing high levels of uncertainty. When estimating the key cost parameters of an inventory model the experience and learning capabilities of the planners affect efficiency of the inventory system. Fuzzy set theory has been used to model inventory parameters subject to uncertainty, where determining uncertain parameters depends upon the subjective opinions of the decision makers. Due to the repetitive nature of inventory planning, the inventory planner has to perform planning tasks repetitively, and consequently, s/he becomes more familiar with the tasks over time. Familiarity with the tasks suggests that learning takes place during inventory planning. Even though the operator’s learning over time may improve his/her efficiency, prior research on fuzzy inventory management completely overlooked the effect of human learning and learning transfer in their models. To close the research gap in this area, this thesis aims to present fuzzy economic order quantity (EOQ) models with backorders, with the objective of formulating the planner’s learning in estimating fuzzy parameters. After a comprehensive and systematic literature review, it was identified that the studies in the literature lacked the empirical evidence on the existence of human learning. Hence, the methodology of the thesis starts with a set of semi-structured interviews with six industrial staff members from different companies. The interviews helped us to gain insights into how human learning is observed in inventory planning under uncertainty. The main themes that emerged from the interviews were summarized as four propositions, which later assisted us in formulating assumptions for the inventory models. Subsequently, through extending an earlier study in the literature, four fuzzy EOQ models with backorders that took account of human learning and forgetting over the planning cycles were developed. The models suggested situations in which the operator applies the acquired knowledge over the cycles in setting the imprecise parameters at the beginning of every planning cycle. The learning ability of the planner was formulated using the log-linear learning curve and the learning curve with the cognitive and motor capabilities of a human being. In order to optimize the models and derive solutions, an optimization algorithm was developed for the first model and applied later throughout the study. Finally, the developed models were examined using both primary and secondary data sets. In the first step, the models were tested using the data obtained from a study in the literature. Next, a case study company in the manufacturing industry was selected and the related data was collected form the inventory system. The models were optimized for the collected data to derive optimal policies for the case study company, highlighting the gap between the current and optimized policies. The results of the study show that learning and forgetting are relevant in inventory management under uncertainty, and that human learning could improve the performance of an inventory system. Incorporating human learning into decision leads to increasing the number of orders, which tends to decrease the batch sizes and increase the maximum amount of inventory