Benchmarking efficiency of children's garment production process using alpha and ILP replayer techniques

Abstract

Recently, the production process of a manufacturer and exporter of children's garment company encountered some problems. Many of clients of the company were consistently complaining about delays in delivery of their orders. This situation was badly and negatively affecting the image and reputation of the company. To solve the problem, a research team was set up. The team conducted couple of interviews with 25 supervisors of the company as well as 4 top-managers (including production manager, general manager, quality assurance manager, and packing manager), who all were unanimously agreeing on the act that the number of the complaint letters and calls from customers were increasing day by day. Unfortunately the company was using a very traditional and out-of-fashion approach to track and trace the placement of each order, since no information system was being used and everything was handled manually following a paper-based approach. The research team spent 4 months to collect, study and investigate all the available data while transferring and transforming the data from printed documents into digital event logs (including timestamp, process type, activity name, resources, and other additional attributes and so on). Thanking to process mining techniques. The researchers were able to overcome and end this undesirable situation. In general, this study is divided into 3 main parts. In the first part, Alpha Miner algorithm, which is a Petri net-based process discovery approach, was applied in order to simulate and visualize the overall workflow of the processes within the garment manufacturer and exporter company. In the second part, ILP Replayer technique, which is a conformance checking approach, was used with the purpose of comparing and replaying the authentic event logs in hand with a standard model (i.e., a master key previously generated by top production managers) in order to investigate the compatibility and compliance of the real data with the pre-defined rules and conditions. The findings showed 15 violations of the rules (i.e., so-called bottlenecks) or undesirable situations which need to be addressed carefully and immediately. The third part of the work emphasizes on offering solutions and suggestions on how the identified problems were solved and undertaken. Eventually, the proposed approach could significantly improve the efficiency of the orders delivery process leading to increased customer satisfaction, customer loyalty and added-value on the company's brand image and reputation.