Improving parts delivery through data aggregation, analysis, and consumption

Abstract

Data analytics and visualization are topics of significant interest in the business and manufacturing communities. This research investigates the hypothesis that, if production floor managers consume properly analyzed data, then their ability to solve problems and prevent production system disruptions improves. This research tests this hypothesis through simulation and a pilot program on Boeing's closet fabrication line and identifies the types of data managers require to improve their operations. The closet fabrication line struggles to complete orders on time, and this problem serves as the central focus for this research. A root cause analysis indicates that issues delivering parts to the closet fabrication line contribute to this problem. Given this issue, this research applies data analysis and visualization tools to facilitate the process improvements required to solve the parts delivery problem. This analysis supports the validity of the initial hypothesis. The results of the discrete event simulation predict an 11% decrease in the time required to fabricate a closet and a 50% decrease in the number of days late the production line delivers closets. The pilot program yields an 11% reduction in build duration and a 32.5% decrease in the duration of the average late completion, while increasing the percentage of complete kits delivered from 39.4% to 80.0%. While the pilot program encompasses a small data set of ten closets, it provides an initial validation of the hypothesis. These results also indicate that information regarding warehouse inventory status, the production queue, and the priority of orders in the queue are valuable data that managers require to improve manufacturing performance.