Fault Tree Analysis: A Path to Improving Quality in Part Stay Protector A Comp

The automotive manufacturing industry is a cornerstone of the Indonesian economy, shaping a fiercely competitive landscape where industry players strive relentlessly to deliver top-quality products [1]. The importance of this endeavour extends beyond economic considerations, as products distinguished by their exceptional quality and competitive pricing not only ensure customer satisfaction [2-6] but also underpin strategies for both national and international competitiveness [7, 8]. The quest for products to seamlessly blend quality and cost-efficiency is an intricate challenge.

While pursuing quality and cost-efficiency pervades the industry, the manufacturing process necessitates a strategic approach to curtail errors within the production line [9, 10]. This strategic focus seeks to produce goods without defects [11, 12] while judiciously selecting affordable, high-quality raw materials. Among the crucial players in this process is the production department, which optimises productivity and quality and relies heavily on skilled labour, well-maintained machinery, and superior raw materials to minimise defect rates [4].

The manufacturing sector, however, is not immune to the risk of producing high-defect items [6, 13-15] that may inadvertently find their way into customers' hands, incurring significant time and financial losses. This predicament assumes a tangible form at BMI, Ltd., where a concerning customer claim arises from YMI, Ltd., a valuable client of BMI, Ltd. Specifically, this claim pertains to the Stay Protector A Comp type K84A component. Stay Protector A Comp type K84A is a part used as a holder for the exhaust, functioning as a connecting part between the exhaust and the exhaust cover.

In this context, the significance of this issue takes centre stage. BMI, Ltd. is compelled to address this matter urgently, necessitating the recall and replacement of older lots with those subjected to rigorous 100% re-sorting. This measure encompasses additional costs and the deployment of personnel for on-site re-sorting at the customer's premises, resulting in time and financial losses. The repercussions extend beyond mere monetary constraints; the trust of YMI, Ltd. teeters on the edge, with the potential for future collaborations hanging in the balance. A continued influx of defective goods risks eroding consumer confidence, jeopardizing YMI, Ltd.'s reliance on BMI, Ltd. for component production.

In a concerted effort to surmount these challenges, BMI, Ltd. has set its sights on curtailing quality costs and preempting customer claims. This study adopts the fault tree analysis method, offering a structured examination of various event combinations stemming from product claims, intending to isolate critical outcomes and mitigate the impact of quality lapses in the manufacturing process [16-18], with a specific focus on the Part Stay Protector A Comp.

3.1 Data analysis

BMI, Ltd. as an industry player in the field of providing automotive products for stamping machine products, has many obstacles in meeting customer needs and satisfaction. One of the obstacles faced is producing goods that are not 100% OK, or it can be said that BMI, Ltd. is still far from the company's target of zero defects in each production activity. Defective goods found at BMI, Ltd cannot be repaired and can be corrected with a product repair process. However, if defective goods are found at the customer, of course, this can reduce the feeling of satisfaction and consumer confidence in BMI, Ltd. As found in BMI, Ltd., within six months, there were several Claim Reports from Customers aimed at BMI, Ltd. due to sending goods that did not comply with specifications/defective products Table 1.

Table 1. Summary of claim customer

Based on Table 1, it can be seen that within six months, BMI, Ltd. obtained 16 Customer Claims for several products that were sent in a defective condition and were found internally by several customers from BMI, Ltd. After getting a Customer Claim History at BMI, Ltd. will then make a bar chart to find out the highest quantity of defective products sent to customers because the defective products sent with the highest quantity have a higher urgency to be used as an evaluation of production activities at BMI, Ltd.

3.2 Defect identification

Figure 1 visually represents a bar chart based on the History Claims table. Based on this, it can be concluded that within six months, the most defective products from customer claims were the Stay Protector A Comp type K84A product produced in the February lot where 726 Pcs were found which were sent to YMI, Ltd. with products that have no centre holes and Berets. Therefore, data collection focused on the product Stay Protector A Comp K84A. YMI, Ltd. is a customer of BMI, Ltd. and certainly desires that the product received or sent by the supplier has specifications that follow the initial production standards. The product sent does not have the slightest defect. However, in the actual situation, a defective product was sent by BMI, Ltd. to YMI, Ltd.

0601.png

Figure 1. Bar diagram of history claim

0602.png

Figure 2. Critical To Quality

Figure 2 shows CTQ Defective conditions found from the sorting results. CTQ is a quality element or characteristic of a product or service that significantly impacts customer satisfaction or meets established quality standards. CTQ factors must be met very well for the product or service to be considered quality by the customer. Identification and understanding of CTQs are essential to improve quality and customer satisfaction.

In determining the manufacture of CTQ based on the findings of defective products from direct sorting results at YMI, Ltd. by sorting operator BMI, Ltd. Next, a Pareto Diagram will be made to determine the cause of the most significant defects in the Stay Protector A Comp product. The data used in analysing the causes of defects in Stay Protector A Comp customer claims is the number of defective products found at YMI, Ltd. Table 2 shows data on defective product findings.

Table 2. Defective product findings data

Based on Table 2., the next step is to calculate the damage value in the form of cumulative defect and map in the Pareto Chart. The result of calculations from the data on the findings of defective products at YMI, Ltd. After calculating the cumulative number of defects, a Pareto diagram is then made to find out more clearly the comparison of the highest number of defects sent by BMI, Ltd. Figure 2 shows the Pareto Chart of Stay Protector A Comp.

0603.png

Figure 3. Pareto chart of defect in Stay Protector A Comp

Based on Figure 3, off-centre holes have the highest frequency of defects, and the cause of their occurrence must be sought. Defects sent with up to 689 Pcs are detrimental to both parties. From the supplier side, of course, when that many defects are found, you have to incur additional costs to send people to do the sorting directly to YMI, Ltd, as well as increasing production costs for product repairs or wasting production costs because products cannot be reused.

3.3 Root cause analysis

In the root cause analysis stage, it is essential to understand how customer claims affect the production process and customer satisfaction. Significant claims not only reduce the level of customer satisfaction but can also disrupt the production flow as the product must be replaced before proceeding to the next stage.

In this context, FTA was used to identify the root cause of the problem. Identifying common factors that could contribute to the problem of off-centre holes was done. Once the common factors are identified, the next step is identifying the essential events underlying the problem. Figure 4 shows the results of the fault tree analysis for the "not-centred hole" problem in the Stay Protector A Comp product. Each basic event and intermediate event is named to facilitate the calculation of their respective weight values. The most significant weight value indicates the leading root cause that must be fixed immediately.

With this approach, FTA was used to analyze the root cause of the "Hole Not Centred" problem in Stay Protector A Comp products in the initial condition before improvement. This analysis provided a solid foundation to identify the key factors affecting this defect so that appropriate improvements could be implemented to improve product quality significantly.

0604.png

Figure 4. Fault tree analysis of the highest defect

Table 3. Probability of events per month for each basic event

Table 3 shows the probability of occurrence of each basic event from the hole nut not centred down to the basic event to find the root cause. The explanation of the basic events obtained is as follows:

1. Basic Event A1 is taken from the last lot of production where this lot was sent to YMI, Ltd. Then it is estimated that the operator does not sample the process as many as two events because the standard check is carried out at the beginning of the middle of the process.
2. Basic Event A2 is taken from the last lot of production where this lot was sent to YMI, Ltd. Then it is estimated that the Operator concentration decreased and did not follow the procedure.
3. Basic Event B1-1: as many as nine of these events are taken from one month of production, there are nine times production, and each cup and tip production is mixed.
4. Basic Event B1-2 is the same as D1; as many as 9 of these events are taken from one month of production, there are nine times production, and each cup and tip product still needs an identity.
5. Basic Event B2: as many as six events according to the number of times the sorting process is carried out before shipping.
6. Basic Event C1 as many as one event because, according to discussions with Quality Control, uneven electrodes occur once every month with a frequency of use of 20,000 Pcs of Product. There were 116,377 products produced using four spot welding, so 116,377 pieces divided by four spot welds obtained 29,904 pcs per spot welding, and the top electrode was replaced every estimated use of up to 20,000 products so that the top electrode for each spot welding was replaced once a month.
7. Basic Event C2, as many as two events where the use of tip spots in January was 8 Pcs with a total production of 157,995 Pcs, and the standard use of tip spots was one Pcs/ 20,000 products. The use of tip spots in February was six pcs with a total production of 116,377 pcs. So, the number of uses for two months is 14 pcs. Moreover, purchase tip spots one time/two months with the number of purchases of 10-12 Pcs so that at the end of the month, there is a possibility that the operator uses a tip spot that is not suitable because there is no tip spot of the required size.

3.4 Minimal cut set of FTA

The next stage is a qualitative analysis with the Classical MCS method. The MCSs of a fault tree Figure 5 are obtained by traditional Boolean algebra. The Classical MCS method handles the negates of non-coherent fault trees with a delete-term operation. The probabilities are calculated utilizing MCSs. Calculating exact probabilities utilizing MCSs is typically impractical, so the Classical MCS method is based on rare event approximation of the unavailability and importance factors. The rare event approximation is adequate when the basic event's failure probabilities are small. The result of MCSs of a fault tree based on mission time for six months shows the greatest unavailability on (B2) Jig Detection Table 4.

0605.png

Figure 5. Fault tree based on minimal cut sets

Table 4. Hole nut-minimal cut sets

3.5 Discussion

Based on Table 4, Jig Detection is the highest weighted value that causes defects because the inspection Jig does not detect an off-centre hole. Improvements are made according to the root cause of the problem. The leading root cause of the problem is that the inspection Jig does not detect an off-centre hole. Improvements made are minimizing errors in the method and improving classification standards for the use of cups and tip spots to minimize repeated errors that will occur in the process method.

Table 5. Improvement on Stay Protector A Comp

Condition
Before Improvement	After Improvement
qqjie_tu_20231109102951.png	qqjie_tu_20231109103006.png
There is no stopper on the Inspection Jig yet	There is already a stopper on the Inspection Jig

Table 6. Comparison improvement on Stay Protector A Comp

No	Before Improvement	After Improvement
1	qqjie_tu_20231109102817.png	qqjie_tu_20231109103023.png
	There is no product list with the specified code yet	A list of codes for each cup and tip spot is made according to the diameter and the product that uses it
2	qqjie_tu_20231109103037.png	qqjie_tu_20231109103103.png
	There is no model code on the body cup and tip spot	The model code is made in accordance with the standard code list that has been made
3	qqjie_tu_20231109103121.png	qqjie_tu_20231109103135.png
	The storage rack does not have an identity according to the model and diameter of the tip spot	Provision of identity on storage racks so that operators can place cups and tips according to classification standards

The root cause of the problem is that the hole is not centred. The inspection Jig cannot detect products with an off-centre hole defect. Based on this, an improvement was made by adding a stopper in the Inspection Jig on the Stay Protector A Comp product so that the Inspection Jig can detect off-centre defects in the future. Table 5 shows the improvements made to Stay Protector A Comp.

Furthermore, improvements were made to the problem of no standard classification for using cups and tip spots. The basic event is placing the cup and tip spot with different diameters mixed, and there is no identity on the cup and tip spot. Even though the percentage value is insignificant, improvements are also made because it has a significant enough number of events. If repairs are made, it minimizes the potential for non-centred defect holes to return Table 6. The results of the improvements that were carried out were controlled and monitored from July to December 2022. These results show a decrease in non-centre claim defect holes from 689 to 0.

The findings of this study identified that the leading root cause of the off-centre hole defect in Stay Protector A Comp products was the Jig Detection system that could not detect products with off-centre hole defects. Similarly, a study in mining [17], identified the problem of workplace accidents and contributing factors. This evidence shows that root cause analysis is a helpful tool to identify the underlying factors of problems in various contexts, be it in production or workplace accidents. This study used FTA for root cause analysis before improvement. Likewise, a study in public transport [23] used fault tree analysis to evaluate factors contributing to road accidents. Studies in Urban Mass Rapid Transit [24] also applied fault tree analysis to generate safety risk models. This evidence shows that FTA is a tool that can be applied in various contexts to identify the root causes of problems.

The results of this study led to improvement efforts by adding a stopper to the Inspection Jig to detect off-centre hole defects. Similarly, the study in risk assessment of fatal accidents [25] identified several fatal causes of fall accidents found in the construction industry and suggested proactive corrective actions to address these issues. This evidence shows that root cause analysis results are often used as a basis for developing better corrective actions. In the context of comparison, the results of this study focus on improvements in the production process and reducing defects. In contrast, those in public transport [23] and Urban Mass Rapid Transit [24] focus more on risk management in public transportation and train operations. Studies in mining [17] evaluated workplace accidents in the context of improved procedures and management, while those in healthcare [22] focused on implementing 5S management in hospitals. This evidence shows that root cause analysis can be applied in a variety of contexts to identify and address specific problems.

In this comparison, root cause analysis is an effective tool for identifying and understanding the root causes of problems in various environments, be it production, transportation, workplace, or hospital management. The implication is that improvements and proactive measures can be taken based on the findings from root cause analysis to improve quality, safety, and efficiency in various sectors.