Improving users’ product acceptability: an approach based on Bayesian networks and a simulated annealing algorithm
提出一种在早期新产品开发阶段评估用户产品可接受性指数的方法，当指数过低时，利用贝叶斯网络推断和模拟退火算法优化选择改进方案，以最小成本提升可接受性。
Developing products that are properly suited to users’ needs and preferences in order to be accepted is one of the main challenges designers and engineers face constantly. Evaluating and improving users’ product acceptability has become an important research question. Many approaches leave the acceptability evaluation question for the last phases of the New Product Development process (NPD), when an almost finished prototype is available and when there is no time left for important modifications. In the early phase of the NPD process, the project managers need models and methods to evaluate the potential acceptability of the new concept and if required, to define actions to improve this concept. In this paper, a method with two main goals is proposed to tackle this problem. Its first goal consists in evaluating an index of users’ product acceptability. When this index is too low, the second goal concerns the optimal selection of the most appropriate actions (improvement scenario) to increase this previously assessed index and to optimise the supplementary cost. As information collected from users in the early phase is subject to uncertainty, the proposed method exploits the inference properties of Bayesian networks making it possible to make useful estimations of the acceptability index. Furthermore, the improvement scenarios are composed of actions that make it possible to improve different criteria composing the users’ acceptability index. The improvement problem is formulated as an optimisation problem to be solved by a simulated annealing algorithm. In order to illustrate its interest, the proposed method is applied to a real case concerning the design of a medical-stocking threading device.