Up until now, it has been a common notion that instrumental process has no predictability(repeatability.) That is, instrumental process is inherently unpredictable and thus inspection task cannot be standardized. In this paper, we show how to implemen...
Up until now, it has been a common notion that instrumental process has no predictability(repeatability.) That is, instrumental process is inherently unpredictable and thus inspection task cannot be standardized. In this paper, we show how to implement cost-effective Ambient Control around glass furnace, reducing the non-conforming rate and the quality variation of its products. It leads to significantly lowered cost and significantly raised predictability of the production process, allowing 4M standardization.
Frequent unpredictable external disturbances compel operators to change setting values of instruments in ad-hoc manner, This is called "Reactive Control." Such control has serious drawbacks. First, monitoring and adjusting setting values require numerous man-hours. Second, it increases COPQ (cost of poor quality.) In this research, we put forth cost- effective control of external disturbances, using Ambient Control(A/C), especially in large-scale glass furnace. Once ambient control is in place, the setting values can be fixed for most of the time, which could enable genuine 4M standardization. In this situation, even in the event that setting values need to be changed (infrequently), it would be a response according to a "Recipe", a set of response plan which has been worked out in advance by analyzing potential predictable situations. As a result, ambient control would lead to stabilization and increased predictability of production process, which allows 4M standardization, and man-hour and nonconforming rate drop, and the lifespan of a furnace increases.
Along with ambient control, we also propose new test method to verify 4M standardization. The instrument like glass furnace is characterized by non-Newtonian fluid dynamics due to high viscosity. The resulting process inside the furnace exhibit significant level of autocorrelation. This makes unsuitable the use of traditional "Control-chart" method. That is why we need a new method.