An investigation into the extended Kanban control system

Abstract

A Kanban Control System (KCS) is a manufacturing/production control system that uses Production Authorization Cards (PAC or also known as Kanbans) to control the Work-In-Process (WIP) of every stage of the Manufacturing Process (MP). It is attached to every finished product, such that once a customer request is received, it is detached and relayed upstream to re-initiate the production process. Thereafter, the finished product is handed over to the customer. In this thesis, different types of KCS were described and categorized according to their operating behaviours. Three significant pull systems, namely Base Stock (BS), Traditional Kanban Control System (TKCS) and Extended Kanban Control System (EKCS) were investigated. The make-up of the EKCS is a hybrid of both the BS and TKCS combined. It was initially proposed to leverage on the strengths of these two systems. However, thus far, all relevant studies only reported on the qualitative aspect of these systems, but none on their quantitative impact. Thus, the purpose of this research was to study the quantitative performance difference of these systems. Specifically, this thesis’s objective was to draw insights from the differences in performance of the EKCS against the BS and TKCS. This study was conducted in two main phases: first, the analyses of Single Product KCS (SP/KCS) was carried out, followed by Multiple Products KCS (MP/KCS). Both studies assumed only a Single Stage/Server (SS). Matlab was used to optimize the systems, and Arena version 12 used to simulate different parameter settings. The performance comparison was benchmarked with KPIs such as Fill Rate, Average Inventory Level and Average Customer Cycle Time. The results showed that EKCS outperforms its predecessors, TKCS and BS in all scenarios. There were four key contributions to this research. Firstly, a method was proposed to determine the optimal size of Base Stock, S*, and Number of Kanbans, K*, in a SS/SP/EKCS, which was never done before. Secondly, this report confirmed that EKCS outperforms TKCS and BS in both single and multiple product scenarios. This performance comparison was ensured through simulations. Third, methods to optimize both Multiple-Product Dedicated (MP/De) and Shared (Sh) EKCS systems were proposed. This also had never been done before. Fourth, this research showed that both De and Sh/EKCS are equivalent. They operate the same way even though their schematics look different. Despite constant praise for EKCS’ performance in the kanban literature, this thesis shows that it outperforms its predecessor, TKCS, only slightly, and only in certain niche scenarios. The worst-performing system turns out to be BS, as it holds a lot of stock in almost all scenarios. Hence, this research has confirmed again that lean or pull production is more effective than push. Current factory floor managers using BS as their production control strategy should consider switching over to the TKCS, or the EKCS in special situations.