Bottleneck identification and alleviation in a blocked serial production line with discrete event simulation: A case study

Abstract

Aiming at the gap between theoretical research and practical application in the production bottleneck field, we apply five bottleneck identification methods in a serial production line in aerospace industry based on discrete event simulation and Plant Simulation software, meanwhile discuss the influence of the bottleneck machine quantity on the system performance. This paper evaluated the practicability, accuracy and limitation of various bottleneck identification methods at the practical level. The results of the bottleneck alleviation manifest that increasing the number of bottleneck machines can effectively improve the system performance, but the more machine quantity, the smaller performance improvement. More importantly, the paper studies the influence mechanism and function relationship of the bottleneck machine quantity on the maximum completion time from an interesting actual phenomenon for the first time. The function obtains the condition that the maximum completion time achieves the minimum. The research and conclusion of this paper have essential reference significance for production guidance and theoretical research, and can also contribute to narrow the gap between theory and application of the production bottleneck field.

Process bottlenecks identification and its root cause analysis using fusion-based clustering and knowledge graph

“…For example, Lei Q et al proposed a constraint and sensitivity analysisbased identification approach for a bottleneck in a job shop [2]. Caesarita Y et al proposed a machine operational bottleneck identification and ranking model in mine operations [5]. Wang S et al focused on energy storage investment requirements and proposed an operational bottleneck identification approach to analyse them [22].…”

“…After obtaining the statistical document vector by BTMDW and the semantic document vector by Doc2vec, the fusion distance based on these two document vectors was applied to Kmeans. The fusion distance between two documents and is defined as a linear combination of and , with a combination coefficient (0 λ 1) shown as formula (5). Firstly, randomly select initial centres of clusters from document collection .…”

“…Calculate the distance between documents and according to the distance function (5). Update cluster centres .…”

“…Therefore, identifying bottlenecks is the key to accelerating the optimisation process management, which aims to improve work quality and efficiency [2]. For example, identifying production bottlenecks in the automotive assembly process can provide manufacturers with essential decision input [3], identifying bottleneck manufacturing resources in the intelligent workshop can provide a decision-making basis for subsequent production management [4], and identifying bottleneck machines can reduce the influence on the maximum completion time [5].…”