Uncertainty analysis, which assesses the impact of uncertainty of input random variables on performance functions, is an important and indispensable component in engineering design under uncertainty. In this paper, a Saddlepoint Approximation based simulation method is proposed to accurately and efficiently estimate the distribution of a response variable. The proposed method combines both simulation and analytical techniques and involves three main steps: (1) sampling on input random variables, (2) approximating the cumulant generating function of the response variable with its first four cumulants, and (3) estimating the cumulative distribution function and probability density function of the response variable using Saddlepoint Approximation. This method provides more computationally efficient solutions than the general Monte Carlo simulation while maintaining high accuracy. The effectiveness of the proposed method is illustrated with a mathematical example and two engineering analysis problems.
Although numerous firms have been shifting toward automated assembly, most still rely on manual assembly when complex assembly operation is required for large-scaled systems. Furthermore, because firms design variants of a system to satisfy diverse customer needs, they may manufacture these system variants in the same assembly line. This type of operation, called mixed model assembly, may improve the utilization of existing manufacturing facilities; however, it may also increase assembly errors due to interchanging geometrically similar parts between system variants. Design for Assembly (DFA) is a design guideline that assists engineers in designing systems that are easier to assemble. However, because DFA guidelines group geometrically similar parts in the same part category, it may be impossible to distinguish geometrically similar but functionally different parts (modules) used in different systems. This paper proposes experimenting how cognitive effects of non-geometric part features influence the productivity and quality in mixed model assembly operations. Furthermore, because the productivity and quality of manual assembly may be influenced by the motivation of operators, this paper examines how productivity and quality may be influenced by different incentive schemes.
In the past decades, firms have increased automated assembly operation to improve productivity and reduce human errors; however, manual assembling is still a necessary operation for complex and large-scaled systems that require high reliability. Furthermore, since customers demand more variety in systems, firms increasingly assemble variants of a system in a single assembly line. In this mixed model assembling operation, there are higher chances of assembly errors due to interchanging of geometrically similar parts between system variants. Design for Assembly (DFA) is a design guideline that assists engineers to design systems that are easier to assemble; however, DFA does not provide any guideline for simultaneously designing variants of system being assembled in mixed model operation. Furthermore, incentive schemes for assembly operators that may influence both assembly productivity and errors have not been the scope of DFA research. In this research, the authors conducted assembling experiments with students to investigate how non-geometric part information and incentive schemes affect the assembly productivity and quality in mixed model assembling operation.
The success of any product in today’s competitive market is dictated by its ability to satisfy the needs of the customers. In this effort, it is important to group similar needs to recognize representative needs, and then identify product requirements that can fulfill these representative needs. One approach to this is to apply Subjective Clustering (SC) to sample data (grouping of customer needs by a sample of customers); however, clusters obtained by SC give only a point estimate of the primary clusters of customer needs by the entire population of customers (population primary clusters). Applying Bootstrap to SC (BS-SC) helps engineers to make inferences on the population primary clusters. In this paper, we randomly pulled out samples of different sizes from both the simulation approach using simulation-generated population data and the empirical approach using experimental population data, and compared the accuracies of SC and BS-SC. Regardless of population sizes, when the sample size was small, BS-SC was more accurate than SC in estimating the population primary clusters. Also, the BS-SC and SC estimates were similar for both simulation and empirical approaches.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.