We applied human-centered design methodologies to enhance the presentation of product quality information to operators on a manufacturing plant floor. First, an initial visual display concept that integrated a pictori...
详细信息
We applied human-centered design methodologies to enhance the presentation of product quality information to operators on a manufacturing plant floor. First, an initial visual display concept that integrated a pictorial representation of a product with standard graphical and tabular information about the product's quality was refined through iterative design and testing. A preliminary study was then conducted to determine the specific features of such a display (termed a pictorial control chart) from among eight candidate detail designs. Finally, a formal study was conducted to compare the performance of operators using this refined pictorial control chart design with their performance using a conventional control chart. Operators completed a quality control task in significantly less time using the pictorial control chart. There were no significant differences in the number of errors committed with the two charts. Subjective measures showed a significant preference for the pictorial control chart. Actual or potential applications of this research include the development of quality control tools that are useful to and usable by operators on the manufacturing plant floor.
This paper addresses the issues of neural network model development and maintenance in the context of a complex task taken from the papermaking industry. In particular, it describes a comparison study of early stoppin...
详细信息
This paper addresses the issues of neural network model development and maintenance in the context of a complex task taken from the papermaking industry. In particular, it describes a comparison study of early stopping techniques and model selection, both to optimise neural network models for generalisation performance. The results presented here show that early stopping via use of a Bayesian model evidence measure is a viable way of optimising performance while also making maximum use of all the data. In addition, they show that ten-fold cross-validation performs well as a model selector and as an estimator of prediction accuracy. These results are important in that they show how neural network models may be optimally trained and selected for highly complex industrial tasks where the data are noisy and limited in number.
暂无评论