Research On Organization Clustering And Process Optimization In NPD Projects

In the current globalization, economic market's competition is increasingly competitive, in order to win the market, new product development(NPD) companys must focus on customer demand and constantly introduce new products of high quality. The NPD project is a complex system architecture which consists of three highly coupled areas "product-process-organization". The analysis of the product's demand is the main factor affecting the product development success. The objective of process optimization is to shorten the time of the project development. And the organization optimization consists of reducing development costs and management complexity, which both ensure the smooth of the product development. So far, the research on these aspects has made a lot of achievements, but there are still some shortcomings in the theoretical methods and practical applications. In this thesis, design structure matrix (DSM) theory and other methods are presented to analyze the characteristics of process and organization of NPD project, the main research contents are as follows:(1)To design and develop a complex product, a firm must answer two critical questions:how to design the right product and how to develop the new product right. This paper provides a big-data oriented method to identify potential customers'behaviors before the product design, and an entropy-based two-stage DSM clustering method to optimize the NPD process.In order to design the right product, this paper uses the PageRank algorithm to investigate customer demands, and so that help the firm select appropriate function modules of the new product; to develop the new product right, we apply the Multi-Domain Matrix (MDM) to identify the interactions among the firm's different departments/units. By proposing an entropy-based method to measure the complexity of PD architecture, which is presented by the product's DSM, a two-stage DSM clustering method is developed to help the firm optimize its PD organization:the first-stage clustering criterion maximizes the added average dependency strength among relevant units, and the second-stage clustering criterion minimizes the weighted total entropy of DSM, including both the entropy of intra-cluster and the entropy of extra cluster. We use a real case of smart phone development to illustrate the proposed models.(2) In NPD projects, coordinating the complex patterns of technical communication among teams is a fundamental challenge. To establish a more effective approach to designing a PD organization for improved coordination, this paper presents anew DSM-based spectral clustering method. Firstly, this paper combines an information connectivity matrix with a responsibility measurement matrix to rank PD teams.Then,in order to measure the similarity of teams, this paper combines the teams'receiver similarity and sender similarity index to derive an integrated similarity matrix. Furthermore, this paper applies the PageRank algorithm to identify core teams in a PD project, which are used as initial seeds for the spectral clustering algorithm based on the integrated similarity matrix to optimize the organization DSM.We compare the optimal results from the proposed method to results from two other approaches to confirm that the proposed approach is a more effective clustering algorithm for reducing PD management coordination complexity.(3)In order to explore the impact of iteration in the NPD process on schedule and cost of projects, this paper analyzes the rework change propagation and builds the rework factor DSM based on the direct and indirect rework probability. The impact of rework change propagation, the time and scope of iteration on rework risk are also analyzed. Then, for reducing the rework impact strength, the method of rework pretreatment is proposed and the model of rework impact after pretreatment is built. Furthermore, the models of project's net value is proposed based on the rework pretreatment and rework risk models. The process DSM optimization objective is maximal the project's net benefit. Finally, an industrial example is provided to illustrate the proposed models. Results indicate that the optimized DSM can reduce the project risk and increase the project's net benefit significantly.