Integrated Decision Support Aid, Multi-objective Evaluation, AHP, Activity-base, IDEF0, Work Flow Simulation
Rapidly changing markets and rapid introduction of new products has created a growing need for agile and responsive manufacturing. Key technologies to reduce time for market and a highly diversified customer base are those of the flexible and reconfigurable decision process. To keep up with global competition, therefore, it is needed to develop a decision support aid that can provide valuable information in a quick, understandable, accurate, and consistent manner. In such problem-solving, the following aspects should be emphasized particularly: (1) a technological goal is usually composed of multiple attributes that are characterized by quite different types of expression, and have incommensurable measures with each other; (2) cooperative work is in nature; (3) dynamic decision environment makes the boundary conditions of the problem very unclear and changeable.
With this point of view, we have proposed an integrated approach for supporting evaluation and review for business decision. To realize the system, we call attention on two important phases for systems development, e.g., value system modeling and functional modeling of a manufacturing process. In practice, we use an activity-based model known as IDEF0 as functional modeling method, and describe the value system model as a tree structure that matches well with a multi-objective optimization method like AHP. We show that being agreeable to the cooperative work, IDEF0 is also helpful for the construction of the objective tree. Moreover, various links with simulation tools enable us to combine the functional model and simulation into an integrated structure.
On the other hand, since the subjective judgement in AHP can apply to any types of attribute, multi-objective evaluation is easy and straightforward even in such a case. For example, simulation carried out here is possible to reveal some features such as productivity, flexibility, and/or operability without the construction of actual plants. Also some statistical aspects may evaluate a certain reliability and risk, and the sensitivity measure is amenable to the evaluation of attributes like aesthetics, comforts empirical items, etc. By giving a subjective measure through the pair comparison or a numerical score to each attribute, we can handle any of those different types of evaluation commonly and evaluate it through lambda-fuzzy computation.
After all, using the proposed approach, we can make systematically and definitely the best decision for the business practice through total evaluation. Moreover, the hierarchical structures both in the value system model and the functional model are very suitable for the modular design. That is to say, model management against the dynamically changing decision environment will be done more quickly and easily, and thus reusability of both models increased.
To demonstrate and validate the approach, a case study has been carried out in an automobile manufacturing company.
System Create Lab. - Research Areas