Two-Step Textual Similarity-Based Approach for Predicting Suitable Production Line for a Newly Designed Product
Year: 2025
Type: article
Abstract: During industry 4.0, digital technology has been integrated with manufacturing processes to improve operational efficiency and hence to enhance organization competitiveness. To this end, computerized methods have been rapidly developed to tackle various production and delivery issues. Production planning and scheduling often demand substantial resources, especially in terms of manpower and time. Consequently, minimizing the time dedicated to both planning and scheduling can hasten product delivery. This paper proposes a novel algorithm that analyzes an unseen process and then predicts a production line by using two-step similarity measures, used in ensemble within the process. Provided with an unseen product model, it identifies the most suitable line, based on the availability of machinery required by the process. In the experiments, eight similarity measures were assessed, based on a realistic production plant. The results revealed that Jaccard similarity and Dice similarity coefficients gave the most accurate predictions. The proposed method is thus believed to be applicable in dynamic production scenarios. Moreover, the developed system also supports incremental production lines.
Source: Engineering Access
Institutions: Burapha University, Suranaree University of Technology
Language: English
Subfield: Industrial and Manufacturing Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
Funder: Burapha University
OpenAlex ID: W7220092339
FWCI:
Cites:
Cited by:
Related to: