Research On The Modeling And Optimization Of Process And Organization In Complex R&D Project

Today, under the driving of technology and market, the function of the new product is getting more and more powerful, and the component is becoming more and more complex. The complexity of the products can better meet the needs of the consumers and brings more business opportunity for the enterprise, however, it makes the research and development process more complicated, which also leads to organization structure and communication between the team become very complex. Therefore, in complex R&D project management, the difficulty not only comes from the design task itself, but also from the management complexity, and how to manage efficiently the complicated relationship between components, process and organization play an important role. Based on Design Structure Matrix (DSM) and Quality Function Deployment (QFD) method, the paper researches on the process and organization modeling and optimization in complex R&D project, the main research contents are as follows:(1) In order to analyze the coupling dependency relationship among functions, products and processes in complex R&D projects, this paper proposes a multi-level waterfall model to analyze dependency in R&D projects based on DSM and QFD. Firstly, we build a QFD matrix representing the relationship between functional and design parameters for quantifying the influence of function modification on design parameters, product and process. And a model of measuring dependency strengths in the function DSM is presented. Secondly, the model deducing a derived DSM matrix from a known DSM is built based on Multi-Domain Matrix (MDM), for example, from function DSM to product DSM, and from product DSM to process DSM.(2) Further, the original process DSM obtained from the above multi-level waterfall method is conducted sequence optimization. The paper adopts the minimized "the total coordination costs" as the objective function, which take into account a variety of factors such as the feedback points, feedback distance, rework probability, rework impact strength, activity cost and duration into consideration, by applying the genetic algorithm and MATLAB software to optimize process DSM. Finally, simulation is conducted to evaluate the optimized effect, and the results show that the optimization method adopted in this paper can effectively reduce the project cost, shorten the project duration and improve project performance.(3) Overlapping between activities is a fundamental feature in R&D project. Even though overlapping can reduce product duration, it also leads to rework risk and increase cost related with overlapping. In order to explore the impact of multi-communication between overlapped activities on rework risk, this paper builds the multi-communication-risk overlapping model and the calculation models of upstream activity’s information modification rate and downstream activity’s rework sensitivity. Then, the models of downstream activity’s rework risk and rework workload are built. Furthermore, maximizing the objective function of saved project cost, the optimization results obtain the best communication frequency and communication interval. Finally, an industrial example is provided to illustrate the proposed models, the optimal results are conducted sensitivity analysis and simulation analysis which yield several managerial insights. Further, based on the analysis between two overlapping activities, process DSM is built, and sequencing optimization and simulation is conducted by the genetic algorithm.(4) In order to analyze the dependency relationship among functions, products, technology measure and teams in complex R&D projects, considering the influence of the technology measures on functions and components, this paper proposes an extended domain mapping matrix (EDMM) model based on traditional DMM. Then the technology measure DSM is deduced according to the relationships among technology measures, functions and components, which can also reflect the dependency relationship between teams. Finally, clustering optimization is conducted and an optimized organization structure is formed whose internal communication is maximized and outer communication is minimized.