Predicting manufacturing production time using machine learning algorithms
No Thumbnail Available
URL
Journal Title
Journal ISSN
Volume Title
School of Business |
Master's thesis
Authors
Date
2023
Department
Major/Subject
Mcode
Degree programme
Information and Service Management (ISM)
Language
en
Pages
101+14
Series
Abstract
Accurate prediction of production time is a paramount challenge for industries, as it directly impacts scheduling, resource allocation, and operational efficiency. This study addresses this challenge by investigating bottlenecks in the daily communication process for predicting production time in manufacturing and proposes a solution through the implementation of Data Automation and Machine Learning techniques. The research aims to assess the effectiveness of various machine learning algorithms in predicting production time, ultimately contributing to improved operational efficiency and resource allocation. To achieve these objectives, the study employs the Design Science Research Methodology (DSR), tailored for IT-based solutions. Two rounds of interviews were conducted to understand the challenges faced by the company and to gather feedback on the implemented machine learning solution. Time study techniques were employed to measure the impact of the machine learning solution on communication processes. The research outcomes demonstrate that a machine learning model, integrated with data from SAP and MES systems, can accurately predict production defect fixing time, thereby enhancing communication processes and overall operational efficiency. The practical implications of this study emphasize the potential of machine learning models, particularly multinomial logistic regression, in improving communication processes within organizations. The integration of machine learning algorithms into existing information systems, such as ERP and MES, is recommended to automate information flow, minimize manual processes, and reduce errors. Additionally, the study underscores the importance of data automation techniques, such as SQL, DAX, Python, and SSIS, to facilitate the integration of machine learning models and data deployment, enabling real-time, informed decision-making. Theoretical implications extend to lean management, value stream mapping (VSM) techniques, machine learning models, and information systems in manufacturing. While the case study project achieved a satisfactory accuracy in predicting fixing time using a multinomial logistic regression model, there are limitations including the need for additional, better data quality to further enhance accuracy and prevent overfitting. This study serves as a milestone, inspiring future data automation development endeavors across various aspects of the business. Overall, this research aims to contribute to the advancement of production time prediction in manufacturing through the synergy of data automation and machine learning techniques.Description
Thesis advisor
Finne, MaxKeywords
machine learning, data automation, information system, manufacturing industry