Title: Defect detection while setting up an assembly line - analytical approach to reduce the N-dimensional solution space

Authors: Malolan Sundararaman

Addresses: Department of Management Studies, National Institute of Technology Tiruchirappalli, Tiruchirappalli – 620015, India

Abstract: While setting up an assembly line, challenge of defect detection is twofold. The first is determination of possible defects and second is examination of a newly setup assembly line to identify the occurrence of possible defects. Here, a two-step analytical approach to address this managerial problem is proposed. In the first step, past setup data are utilised to identify: 1) list of possible defects; 2) list of control parameters, along with the possible operating levels, to construct an N-dimensional solution space. A bootstrapping approach generates bootstrapped sampling distributions for each sub-solution space. From this, probability of occurrence for a specific defect when control parameters are operated at a particular level is identified. In the second step, an optimisation model determines the ideal sub-solution space, ensuring all defects are identified with an acceptable level of certainty. Data from a developed experimental design is used to demonstrate the proposed approach empirically.

Keywords: defect detection; assembly line; bootstrapping; optimisation; probability mass function; PMF.

DOI: 10.1504/IJAOM.2025.145209

International Journal of Advanced Operations Management, 2025 Vol.16 No.1, pp.109 - 134

Accepted: 04 Oct 2024
Published online: 25 Mar 2025 *