Vol.5 No.1 2012
9/82

Research paper : Construction of a traceability matrix for high quality project management (A. Sakaedani et al.)−6−Synthesiology - English edition Vol.5 No.1 (2012) quantification of difficulty is made possible by evaluating the magnitude of the entire system matrix n. Therefore, the evaluation will be done according to how many times the matrix is greater than the unit matrix, or the Euclidean norm of system matrix n. However, the component value of system matrix n with respect to difficulty is determined by evaluating the difficulty of the elements (guideline for difficulty setting: Reference Material 1). The standard value is 1, and the value becomes greater than 1 at higher difficulty, or less than 1 at lower difficulty.4.3 Definition of complexityThe complexity of information transfer is defined as follows:Complexity = difficulty × interdependencyHowever, difficulty is given by the Euclidean norm of system matrix n, and interdependency is given by the Euclidean norm of system matrix s minus the unit matrix (sample calculations: Reference Material 2).The difficulty and interdependency are variables that reflect the situation of the individual elements and relationships among elements of the project. Therefore, by understanding their indices, it is possible to understand the elements of the matrix and the relationships among them. At the same time, the whole project can be understood via the changes in complexity obtained by multiplying indices. 4.4 Obtaining a square matrixThe relationships among the elements are not necessarily in a square matrix. In such a case, it is necessary to form the square matrix for the diagonalization that is necessary for the calculation of interdependency. To form the square matrix, the component value 0 is given. Since the component value is 0, the interdependency value changes as much as the degree added to the square matrix formation (this is because the diagonal component is subtracted for the row or column added by the square matrix formation, and the additional diagonal component becomes -1). However, considering that complete independence is expressed by the square matrix called the unit matrix, the change in the interdependency value by adding the component value 0 in the non-square matrix must be understood as the index that indicates the independence of the non-square matrix. From the above, the square matrix formation by adding the component value 0 is set as the rule.5 PDCA cycle of project management using the traceability matrixThis chapter explains the methodology for understanding the whole of a project, and its details, and for “seeing both the forest and the trees.” For explanation, the PDCA (plan do check act) cycle will be used as the scenario (Fig. 4).First, to create the traceability matrix, specific elements are organized and the matrix is created. The complexity of the project as a whole is reduced by improving the difficulty and interdependency in the created traceability matrix. The actual development activity is conducted in the project, and the progress is checked. The complexity of the project is used as the index of progress, and an overview of the state of the project is gained by looking at the change. If the change of complexity is on an increasing trend, it is because some problems have developed in the project, and the causative element is sought. The difficulty and interdependency of the elements are changed by simulation, and the element that Fig. 4 PDCA cycle of project managementStartExecute projectExecute measures for highly sensitive elementsChange the difficulty of elements and analyze the sensitivity of complexityIs the rate of change positive?Check the rate of change of complexityConduct reviews at appropriate intervals (e.g., weeks or months) by each procedureEvaluate complexityReduce difficultyReduce interdependencyAre the granularities of elements varied ?Can the relationships be organized ?Organize issues of project plan(organization is vague, activity flow is vague, etc.)Create system matrixUncover elements and organize relationshipsPLANCHECKACTDONoYesNoYes

元のページ 

page 9

※このページを正しく表示するにはFlashPlayer10.2以上が必要です