Control diagram of depth data to monitor multivariat process.

Abstract

In this more competitive globalization era, business performers will pay more attention to the quality to win the competition. Understanding variation and how to reduce it is the key  in improving the quality. Multivariat quality control in production process not only focus on minimal control but also control of variability process. The methods that can detect multivariat cases such as T2 Hotelling, MEWMA and so on have weakness because they totally depend on normal distribution. In fact, not all processes can satisfy normal distribution. This study describes control diagram of depth data by using multivariat case depth of as mahalanobis as an alternative because it doesnot need any assumption about distribution underlying process. Its implementation done by analyzing satisfaction data of clients in a hotel in Cihampelas on standard rooms. The results show that the abnormal distribution is in 10%. Satisfaction control of the clients using control diagram of depth data shows that the process is out of control. Displacement estimation happens on 21st period. When control diagram of T2 Hotelling is used, the process is in control. Therefore, when the normality assumption is broken, and there is parameter displacement of location. The control diagram of T2 Hotelling doesnot immediately provide the signal of  out of control.