A comparative study on kanban, CONWIP and hybrid systems with illustration of a case study
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Production is a game of producing more with less, which can help a company to keep competitive in the market. The traditional batch-and-queue type of manufacturing or push system has been challenged by pull systems, such as CONWIP and Kanban systems.
This thesis explains in detail the difference between Push and Pull production systems and, the difference between CONWIP and Kanban systems in application.
A real-world case study is conducted to help identify the best system suited to the production line through simulation by rating its throughput rate, Work-In-Process level, and card number in system. Three types of production systems are developed and modeled: (1) Kanban system. (2) CONWIP system. (3) Hybrid systems, which integrates CONWIP and Kanban in one system.
Based on the simulation results, the following conclusions have been made: (1) For Kanban and CONWIP systems, with the increase of the number of cards in system, the output does not necessarily increase; it stays the same once saturated, while the WIP increases. (2) For CONWIP system, the location of the bottleneck does not have much impact on throughput or on WIP level. (3) Most of the time, CONWIP outperforms Kanban system by being able to reach desired (or maximum) throughput with fewer cards in the system to manage than Kanban. (4) A hybrid system with a CONWIP system managing the first station through the bottleneck and a separate CONWIP system managing the rest stations performs better than or equal to pure Kanban or CONWIP systems. (5) The number of needed cards in the system can be determined through Little's Law or tried out with the guidance of critical WIP. (6) A well designed hybrid system will perform as good as or better than a pure Kanban or CONWIP system.
This research also shows that simulation modeling is very useful for production professionals to optimize their production system parameters and open opportunities for future research.