Implementation of Statistical Process Control through PDCA Cycle to Improve Potential Capability Index of Drop Impact Resistance: A Case Study at Aluminum Beverage and Beer Cans Manufacturing Industry in Indonesia

Purpose: The purposes of this study are first, to analyze why the process capability index (Cpk) for drop impact resistance (DIR) does not meet the specification or less than 1.33, and second, to find out what improvements should be made to make it meet the specification. Methodology/Approach: The methodology used was Statistical Process Control (SPC) through the PDCA cycle, supporting with Cause and Effect Diagram (CED), Nominal Group Technique (NGT) and “why, what, where, when and how (5W1H)” method. Findings: With the above methods, the result of the study was given a positive impact on the company. The average of DIR was increased from 20.40 cm to 25.76 cm, increased by 26.27% and the standard deviation was reduced from 1.80 to 1.48, and then the Cpk index was increased from 0.48 to 1.79 it means the process is in control and capable. Research Limitation/implication: This research was limited only on the twopiece can aluminum cans manufacturing process, no for three-piece cans manufacturing. SPC through PDCA cycle is an interesting method for continuous improvement of process capability in the cans manufacturing industry. Originality/Value of paper: This study highlights the area of future research SPC through the PDCA cycle to analyze and optimize process capability. Therefore, this research is considered to promote and adopt high-valued methodologies for supporting industry to achieve global competitive advantages. Category: Case study


INTRODUCTION
The Canmaker Magazine Vol 32: February 2019 reported that the beverages cans demand in the Southern East Asia region is about 7.2% from the total 335 billion of the global beverages cans demand. Aluminum cans have experienced many important developments throughout the years, if compared with other packaging aluminum cans having some advantages such as good in the stackable, easy opening by full the tab, hermetic sealing, environmental and economic (Mohamed, 2016).
With all these advantages and a good trend in the market, that condition redirects in progressively savage challenges to get each other's chances. Aluminum beverages cans manufacturing industry located in Jakarta is one of the packaging industry in Indonesia that has engaged with the challenge to become the cost leader and also to remain competitive in the global world packaging industry today, with slightly process changes in tooling geometry of punch sleeve and activated oven washer dryer zone 2 to achieve the minimum requirement of the potential capability to be able to produce the aluminum beverages cans with the new aluminum raw materials (Y1) without any reducing or degradation of the aluminum cans product quality.
The fact was with Y1 material; from nine critical parameters, there is anyone of them the potential capability index (Cpk) does not meet the minimum requirement, the parameter is DIR with the Cpk index achievement was 0.48. The investigation intends to get the reasons for the faulty. Statistical Process Control (SPC) through the PDCA cycle and Nominal Group Technique (NGT) are combining in utilized to find out the root cause and the 5W1H method was used for improvement determinations. As one of the facts in the real industry that the defected of the products can be reduced effectively by the Integrating of nominal group technique, Shainin system, and DMAIC methods (Trimarjoko et al., 2019).
Quality improvement is becoming a critical issue in the highly competitive business environment nowadays, so the products are made need to be upgraded regularly (Dhounchak and Biban, 2017). Minimizing defects during the process is one thing that needed to maintain customer loyalty (Realyvásquez-Vargas et al., 2018). To make high-quality products proper planning and preparation are categorized as a vital factor (Chakraborty, 2016). The organization needs to maintain a process and keeping continuous improvement to make good product quality and minimize defects (Nugroho, Marwanto and Hasibuan, 2017). To reduce defects and minimize process variation can be used in the DMAIC method (Ani, Ishak and Shahrul, 2016). The investigation of a problem and the development of an appropriate solution to the quality improvement process could be able using quality tools (Nabiilah, Hamedon and Faiz, 2018). Many statistical tools are available to improve process; one of them is SPC (Statistical Process Control). SPC could be able to maintain process stability and capability (Saputra et al., 2019). The deviation or variation of the process can be eliminated, and also, the process capability can be increased by applying the DMAIC method (Sharma and Rao, 2013).
In the competitive business environment, the SPC method could be able to use to improve the process (Godina, Matias and Azevedo, 2016). Claim from customer need to be controlled well, SPC is one of the many tools it's effective in reducing claim (Solihudin and Kusumah, 2017). Quality could be able to increase by solving problem happened during the process, and the SPC method can be used as tools (Devani and Wahyuni, 2017). Product quality must be maintained; it can be implemented by building the team to aware of SPC (Mangesha, Singh and Amedie, 2013). Product quality control with the SPC method helps in reducing defects (Supriyadi, 2018). A process could be evaluated effectively by seeing the capability process index, which can be used as a managerial decision (Sagbas, 2009). Identifying the critical to the quality of a machining and prioritization corrective action are strongly needed for the improvement step, and the DMAIC method can be used to increase capability index level (Sharma, Rao and Babu, 2018). Quality satisfaction can be created through statistical tool implementations like SPC and FMEA (Rana, Zhang and Akher, 2018). Control chart standard deviation (S) and (X )) are a statistical tool which can be used to a created quality products, and it could be able to upgrade to becoming a highquality product with controlling the range and capability process coefficient as the indicator (Fazeli and Sharifi, 2011). To create a quality of the product are made has to be started from a small issue which was happened on the process or organization (Tuna, 2018). The quality is essential for the product that was made. It can be maintained with SPC implementation as robust tools (Bereman et al., 2014). Quality could be improved by emphasizing all the levels in the company to discipline to use statistical tools (Sokovi, 2009). The organization or company need to be aware to avoid mistake or wrong in doing an operation. QC with 7 tools is a switchable tool (Magar and Shinde, 2014). Also, in anticipating globalization, the product defect must be reduced, Kaizen and PDCA cycle are a famous tool to make it happened (Darmawan, Hasibuan and Hardi Purba, 2018). The organization or company that have many product types, a lot of checked quality parameter and also the materials came from many sources it recommended to implement the assessment process with monitoring stability and capability (Ramirez and Runger, 2006).

RESEARCH METHODOLOGY
The study aims to find out the factors were causing potential capability index (Cpk) of DIR for aluminum beverages cans does not meet the customer requirement or common industrial standard and constructing the steps of corrective to improve it in minimum 1.33. The conceptual frameworks for this matter are illustrated in Appendix 1 ( Figure A1).
Based on research framework as in Figure 1, for getting potential capability index (Cpk) is meeting to the customer standard, SPC through PDCA cycle with the integration of NGT and 5WIH methods are used, supported with some of the statistical tools such as (X,) ̅ R chart, histogram, and fishbone diagram.

RESULT AND DISCUSSION
Process improvements must be given high priority and documented. By using SPC through the PDCA cycle followed with CED and NGT to identify the root cause, then continued with 5W1H methods for determining improvements, the steps to achieve the above matter as in Figure 1.

Steps/cycle Activity Plan
Data collection, determine research priority and interview.

D0
Making a plan and do an improvement with the 5W1H method.

Check
Stability process ( -R Chart) and process capability (Cpk).

Action
Making standardization.

Determining Improvement Priority
The capability study data, as in Table 1, concerning the potential capability index (Cpk), plotting to the trend chart to get easier in the analysis, as shown in Figure 2.

Figure 2 -Trend chart of Potential Capability Index (Cpk) Aluminum Beverages Cans With New Aluminum Raw Materials
Showing up Figure 2 above clearly that the DIR parameter is needed to be improved due to the achievement of the potential capability index (Cpk) was less than 1.33.

Potential Capability Index (Cpk) Aluminum Beverage Cans After Changed Aluminum Material Input
Cpk LSL

Discussion in Determining The Root Cause
The discussion was done with the staff of packaging aluminum beverage cans plant in Indonesia consisting of production, corporate production, and the Quality Assurance department. The aim of the discussion is for getting optimum results in solutions. Table 2 is describing the result of the discussion or brainstorming regarding the possibility of the root cause for the faulty drop impact resistance capability with the index less than 1.33. By observing Table 2, from 17 items of possibilities were causing for DIR does not meet to the customer specification in term of the Cpk achievement, to make clear in analysis the next table will be given classification information in more specific and details as stratification. Table 3 is describing the cause of the human (man) factor, the cause of the material factor as in Table 4, the cause of the method factor is in Table 5, the cause of machine factor is in Table 6, the cause of the tooling geometry factor is in Table 7, and the cause of the environment factor is in Table 8.     From the above stratification data in (Table 3-8), the next step is plotting into the cause and effect diagram (CED) with the aim to determining the root cause of why the potential capability index of drop impact resistance parameter for aluminum beverages cans does not meet to the customer requirement or common industry standard.

Creating the CED
CED to determining the possible root cause, as shown in Figure 3.
Analyzed CED as in Figure 3, there were ten (10) the possible root cause of the potential capability index (Cpk) does not meet to the customer requirement as can be seen on the rectangular box with dashed lines, details of the possible root cause are as follows: (1) Man: The possibility of the operator did wrong or a mistake in measurement and lack of knowledge.
(2) Material: Yield strength and thickness (3) Method: Air pressure that injected inside the cans and base plate thickness.
(5) Tooling: Punch sleeve tooling geometry and Aluminum thickness at dome area.
Based on the six factors above with ten findings cause were considered as the potential sources of the cause.

Creating the NGT
The next step is determining what the dominant cause for the issue. NGT method was used, the discussion group with eight members to involve in giving the score for NGT. All the members were coming from different backgrounds such as education, age, year of service, and current expertise. With these differences in the various background, it will be resulted in more accurate in giving the information, and finally, the correct decision is gotten. The concept of it is as in Table 9.  Interpreted of Table 9 linked to the NGT value with using equasion (1) there are three (3) potential variable causes have NGT higher or the same 41, that became a dominant factor of the cause for the Cpk DIR achievement, namely: Air pressure injected to inside the cans, temperature of oven dryer and tooling geometry of punch sleeve.

Making Improvements
After the dominant cause or a vital factor is found, the next step is to determine the improvement steps.

Quality Improvement Plan (Plan)
By identifying the underlying causes, clarifying why they need to be improved, what improvements are being made, where or what areas are being corrected, when actions will be taken, who will improve them, and how to improve them, it will be more targeted. More details, improvement plans with the concept 5W1H are described as in Table 10.

Air pressure is injected inside the cans before the test of DIR
To avoid air pressure that injected inside the cans before test DIR, the pressure gauge indicator changed to the new one and doing the routine check for the function of pressure gauge with monthly bases and put on the calibration schedule.
Temperature oven washer dryer too high > 420 o F To make oven washer dryer temperature does not exceed 420 o F the action has been done is activated oven dryer zone 2. Detail temperature with activated zone 2 is, as shown in Table 11. In line with the above data on Table 11, by activating two oven zones on the typical washer oven dryer, the oven temperature setpoint can be able to set to 395 o F for two minutes and fifty-two secs. The drying effectiveness has been still good and the particular big impact on the lightweight aluminum is getting safer to avoid annealing or softening if the machine stoppage or perhaps jam for a while. Figure 4 below is illustrated typical of oven washer dryer zone 1 and zone 2.

Figure 4 -Washer and Oven Dryer
Tooling Geometry or Profile Punch Sleeve Stretching aluminum at dome area point 3 (p3), as shown in Figure 5 was reached 3.83% from the original thickness, it resulted in the drop impact resistance became weak. p2 p3 p4 p5 Figure 6 below is describing punch sleeve nose radius schematic radius 1 (R1) and radius 2 (R2). Changing details of punch sleeve tooling geometry or profile are as shown in Table 12.

Constructing ' and R Chart
X and R Chart was used to control process stability with the final purposes is to minimize process variations. The below data as in Table 13 is capability study data for 5 hours running after improvement was done.
The sampling was carried out for 5 hours, from 08:15 a.m. to 12:15 p.m, followed by the DIR test with the air pressure were injected inside the cans continuously controlled at 60 Psi, the results of the test as shown in Table 13 above. From these data, we do the calculation to find the central point or Center Line (CL), Upper Control Line (UCL), the carry control point or Lower Control Limit (LCL) and its process capability index or Index Capability Process (Cpk).

The specification of DIR
The specification of DIR is minimum of 17.78 cm

Determining the Centre Line (CL), Upper Control Limit (UCL), and Lower Control Limit (LCL) for X-Chart
Centre Line (CL): The constant for A2, D3 dan D4 for subgroup number 6 is as in Table 14. The above calculation on Eq.

Figure 7 --R Chart Drop Impact Resistance After Improvement of the 3 Dominants Factor
From the control chart in Figure 7, the process statistically was stable; the indication is there was no point is out from the control limit.

Calculating the Potential Capability Index (Cpk)
To calculate the potential capability index firstly need to know the standard deviation, due to the capability index, was decided using Cpk on this research so the standard deviation directly can be calculated as estimated standard deviation.
The estimated standard deviation could be calculated using the below equation (8).
Determining the standard deviation (s): DIR has only had one side specification (minimum specification), and the Cpk = Minimum (Cpu, Cpl), due to only one side specification so the Cpk will be the same with Cpl (6;< = 6;=). To determining the Cpk, the formula used as on (9). The Cpk calculation based on the equation (9) is plotted to the histogram using statistical software NWA Analysis v6.3 with the result as in Figure 8.

Figure 8 -Histogram of Drop Impact Resistance after Improvement
The histogram in Figure 8, given the information that the potential capability index of DIR after doing action on the three (3) factors that causing faulty in the potential index capability achievement, is positive with the Cpk index 1.79.

STANDARDIZATIONS
Looking at Figure 9 below we can learn that after improvement the average DIR was increased from 20.40 cm to 25.76 cm, and the potential process capability index (Cpk) increased from 0.48 to 1.79.

Figure 9 -The Achievement of Cpk and Average of DIR Test before Vs. after Improvement
The achievement as in Figure 9, then plotted into the distribution plot, as shown in Figure 10.  Based on Figure (7, 8, 9, and 10), the process stability and capability results for the DIR are ideal categorized, and it can be seen in the control and capable matrix in Figure 11. After getting the improvement results, then determining the standardization to maintain the stability and the capability of the process, as shown in Table 14.

CONCLUSION AND RECOMMENDATION
The study was implementing the SPC to analyze the data, PDCA to continuous improvement, CED to determine the root cause and NGT to determine the dominant cause factors, and then 5W1H method to manage the improvement. It's very useful and effective in creating and improves aluminum cans packaging product quality.
The fact is the average of DIR increased from 20.40 cm to 25.76 cm, the standard deviation was reduced from 1.80 to 1.48 and the potential process capability index (Cpk) increased from 0.48 to 1.79, it can be concluded that the process is stable and capable. The significant impact for the company was the company to be able to use aluminum material Y1 to produce aluminum cans packaging with high-quality standards.
To maintain the process stability and capability are always meet to the specification, it needs to be well controlled for the parameters i.e.: 1) Air pressure which is injected into the inside of the cans when doing the DIR test to make sure stables.
2) The washer-dryer oven machine the temperature setpoint needs to be controlled to do not exceed 420° F with two zones oven activation to avoid aluminum softening or annealing if any machine stops for a while. 3) To avoid the stretching during the doming process, which is caused by tooling geometry of the punch sleeve nose radius, it needed to do regular checks to maintain the clearance is match with the Y1 materials.
At the end of this study, further discussion is needed to maintain what has been successfully achieved. The recommendations for future researchers to make it better is highly recommended to use the FMEA method because it has an RPN (Risk Potential Number) index, so it will be more accurate to make justifications establish the improvements. Figure A1 -Research Framework