Product Quality Control with Six Sigma and Preventive Maintenance

PT. X is a garment company in Cimahi, West Java which consists of four departments, namely texturizing, twisting, weaving and dyeing printing. Research focused weaving department with identifying the cause of the defect with six sigma method Define, Measure, Analyze, and Improve. Based on the identification issue using the steps define, measure and analyze known that the causes of defective products is declining performance nozzle and pump components resulting in less water pressure or excess. In addition, not component replacement for a worn shedding motion so that a malfunction of kamran. Therefore, the purpose of this study was to schedule preventive maintenance on spare nozzle, pump and kamran. Improve these problems by creating a preventive maintenance schedule by age replacement method. Based on the results obtained scheduling data processing component replacement on the machine water jet loom is; (1) The components of the pump will be replaced every 150 days, (2) Components kamran 71 days and (3) Components nozzle 82 days.


Introduction
PT. X is a garment company in Cimahi, West Java which consists of four departments, namely texturizing, twisting, weaving and dyeing printing. Department is a texture creation texturizing yarn, twisting the yarn preparation department is spin yarn from being hairy in the knitting process. Department of weaving is a weaving manufacture of semi-finished fabric (gray cloth) and dyeing is a dyeing and printing process is a process where the tasting. This research focused in weaving department with identifying the cause of the defect with the six sigma method and above it.

Define
PT. X producing various kinds of gray fabric in the weaving department uses water jet loom machine. Process water jet loom machine contained five movements, namely shedding, picking, weft insertion, beating up, letting off, and taking up. During 2018 the production of gray cloth are the number of defects is above the tolerance limit of the company that is defective feed neps. Neps feed defect is their arch/loop weft threads in the fabric caused by the launch of the feed is not perfect. Weft threads are threads which are arranged horizontally straight on a fabric and yarn warp yarns are arranged vertically straight while the yarn. These two threads are crossed with each other so as to form a woven fabric.

Analyze
In the analyze phase is carried out to analyze the type of feed neps approach analyzerie FMEA (Failure Mode and Effect Analysis). FMEA result data obtained from questionnaires to the head of maintenance, production and supervisors are presented table 1 below; Based on identification in Table 1 are major causes of disability acquired neps feed. The cause of most dominant factor is the engine in the absence of preventive maintenance, so that the damage caused by an emergent resulting engine of defects. Based on the results of the highest RPN value weft fabric neps defects caused by performance engine components and pump nozzle declines and shedding motion component wear. This happens due to lack of maintenance and replacement of components which are scheduled to cause excessive water pressure or less as well as damage to kamran function. So the purpose of this study is to make a schedule repair or replacement of components of the nozzle, pump and kamran.

Six Sigma
According to Schroeder et al [1] Six sigma accordance with the meaning, is the distribution (variation) of the average (mean) a process or procedure. Six sigma is applied to minimize the variation (sigma). Six Sigma as a measurement system using Defect per Million Opportunities (DPMO) as the unit of measurement. DPMO is a good measure for the quality of the product or process, for directly correlated with disabilities, costs and time wasted.
According [2], the implementation of Six Sigma quality improvement consists of five steps using the DMAIC or Define, Measure, Analyze, Improve and Control.
Define DMAIC is the initial stage in which the stage is done is set a target of six sigma quality improvement activities. In the define phase of defining the characteristics of the system performed as well as an action plan that will be done to process upgrade. Measure is a logical follow-up to define stepsand is a bridge to the next step.
Analyze is a third operational steps in six sigma quality improvement programs. There are several things that must be done at this stage are: (1) determine the stability and capability (capability) of key processes, (2) Establish performance targets of the characteristics of quality (CTQ) and (3) identify sources and root causes of quality problems. Improve In this step implemented an action plan to implement the Six Sigma quality improvement. The plan describes the resource allocation and priority or alternatives do. Six sigma quality improvement team must decide the target to be achieved, why the action plan is carried out, whereby the action plan will be carried out, Control is the last operational phase in efforts to improve quality based on Six Sigma.
Failure Mode and Effect Analysis (FMEA) is a structured procedure to identify and prevent as much as possible modes of failure. A failure mode is any one that is included in defects or failures in the design, conditions beyond the limits established specifications, or changes in the product that causes disruption of the function of the product.
There are three components in defining the priorities of failure, namely:  Severity (Sev) is a bad influence or risk will be borne by the customer as a result of failure caused. Rating assessment provided from the lowest value to the highest value of 1 to 10, where 10 is the worst impacts.  Occurrence (Occ) is the possibility that the cause is going to occur and result in the form of failure during use of the product. With estimated that a possibility of occurrence on a scale of 1 to 10.
 Detection (Det) is a measurement of the ability or failures control that may occur. RPN (Risk Priority Number) is a mathematical formula that is composed of a set of effects (severity), the possibility that the cause will cause failures related to the effects (occurrence) and the ability to detect a failure before it reaches the consumer (detection), The calculation formula RPN defined with the following formula: RPN Occ Sev x = x Det (3) RPN value is used to identify the most serious risks that lead to corrective action. In the FMEA process, the type of failure that has the highest RPN value will be evaluated first to determine a control plan that aims to eliminate or reduce the effects of such failures.

Maintenance Management
Treatment as that activity or component/system that is broken will be restored or repaired within a certain condition in a certain period. According [3]. Care is closely related to the two most important issues are the activities of damage preventive maintenance and corective maintenance. In the process of the action consists of several activities, such as: Preventive maintenance is the activities maintenance and care taken to prevent damage to the unexpected and find conditions or circumstances that can lead to the production facilities were damaged when used in the production process. The purpose of preventive maintenance is directed to maximize availability and minimize the cost of an increase in reliability, with the scope of activities could only cover the area and expanded to other areas and public facilities. The activities conducted in preventive care are as follows; Replaced by new and Imperfect preventive maintenance Corrective maintenance is the maintenance and care activities performed after the occurrence of a damage to facilities or equipment that result can not function properly. This maintenance activity is not scheduled, depending on the condition of a component or system itself. Reliability is the probability that a component/system will inform a function that takes a period of time when used in operating conditions According [3]. In general, the concept of reliability can be described in a Bathtub Curve that explains the life cycle of items/components.  Figure 1. Bathtub Curve Reliability is the probability that the equipment is working satisfactorily by function over a longer period of time and under certain conditions. Statistically ata reliability function R (t) is defined as the probability of a component is still functioning in the specification at the time t, can be expressed mathematically as follows: Where x is the age of the components. If f(x) states and the probability density function f(x) dx represents the probability of a component will and the number of the probability function and cumulative distribution function of reliability equal to one, then obtained the following equation: So that the probability to be operating in the specification or reliability function can be expressed mathematically as follows: The rate of damage to a machine or component at time t is the magnitude of the probability that the machine or components will be damaged in the next interval. While at the time t the machine or machine component dala good condition. Mathematically the damage rate function can be described as follows: If r(t) increases with time, the function of the rate of damage to the ascending and vice versa, if r(t) decreases with time so as a damage rate decreased.

Distribution of Damage
Basically, there are several forms of distribution of the damage that can be used in treatment policies, such as Exponential distribution, Weibul, Normal and Log Normal. The distributions have a constant decay rate is also called exponential probability distribution. The exponential distribution is an important distribution on distribution reliability. Another distribution that can be used is weibull distribution, normal and log-normal. The third distribution function of damage rate is not constant so that it provides an alternative that can be used in addition to the distribution of exponential decay that already exist.
Exponential distribution has a constant rate of damage, do not depend on time. Thus, the probability of occurrence of damage to a part or tool does not depend on the age of the equipment. Weibull distribution have decreased the rate of damage to <1, the rate constant for the damage  = 1 and the rate of damage rise to  > 1. Normal distribution has a rate of damage being up since age tools, which means that the probability of damage to the equipment or component rises with age components. The normal distribution has two parameters, namely the average and standard deviation.
Lognormal distribution knows two parameters: s as a parameter tmed shape and location is the middle value of the time of the crash. This distribution is defined only for positive t value, and therefore better suited as a damage distribution. Lognormal distribution has some form and according to According to Jardine [3], is often also present data in accordance with the distribution weibull. Damage Identification Distributionis to show through the test statistic in terms of accepting or rejecting a hypothesis that damage or repairs under study originated from a specific distribution. Identification of distribution can be done in two phases, namely Index of Fit (r) and Goodness of Fit Test.