Abstract:
Inventory decisions have become more challenging in the context of the coronavirus crisis. The pandemic impacted all of the economic domains. Companies experience the highest level of uncertainty about demand, lead-time, and supply. The consumer's behaviour has changed. The transportation services also plummeted during this period. Moreover, the availability of some raw materials has decreased. The production process was disrupted, and the workforce planning was strictly regulated. In this new environment marred by uncertainty, the decision-makers need to use efficient and quick inventory techniques to maintain optimal inventory levels and ensure accurate inventory control. These inventory techniques should consider the uncertain or vague inventory variables and adapt quickly to the market changes. This paper provides a new ABC classification that deals efficiently with vague criteria and quickly adapts to the changes from the business environment. This hybrid inventory technique integrates the Analytical Hierarchy Process (AHP) and Fuzzy Logic Systems (FLS) in traditional ABC Classification. It employs multiple criteria, such as annual usage, annual usage cost, lead time, and criticality. The AHP helps the inventory managers to rank criteria considering the Relative Importance Matrix. The FLS allows managers to identify the relationships between the criteria and those three classes of ABC traditional classification to classify the company's items accordingly. The integration of fuzzy systems also helps managers to overcome the major limitations of the traditional ABC classification, such as using only financial criteria and modelling only crisp or measurable criteria. The analysis of the hybrid ABC Classification provides a comparative approach to three mixed classifications that combine the traditional classification with Analytical Hierarchy Process (AHP), Linear Programming (LP), and Fuzzy C-Means Algorithm (FCM).
The new inventory classification has both strengths and weaknesses. One strength is that the Fuzzy Logic Systems that classify the items can be modelled easily using software applications or platforms. Secondly, the inventory problem can be solved even if the parameters are vague, while the inventory problem can be adapted whenever necessary. The major weakness pertains to the fact that the entire modelling process is based on the AHP step, where the criteria are ranked only by the decision-maker's reasoning, which may lead to costly human errors. This AHP step in modelling Fuzzy Logic Systems can be replaced with Linear Programming Algorithms to obtain an objective ranking of criteria.
