Performance Enhancement in Injection Moulding Manufacturing: A Case Study Utilizing the DMAIC Methodology

Achieving high productivity without compromising product quality is a central objective for modern manufacturing industries [1, 2]. This objective compels companies to improve process efficiency in order to remain competitive, while simultaneously reducing reliance on imported equipment and components. In Indonesia, the injection moulding sector is experiencing increased demand, particularly for the production of medical devices.

Despite this growth, manufacturers often face significant operational challenges that hinder their ability to meet production targets. One such case is PT. XYZ, a local manufacturing company specializing in medical equipment using injection moulding technology. Based on internal production data, the company's average Overall Equipment Effectiveness (OEE) over the past year stands at 81.72%, which falls short of the company’s performance target of 85%.

Preliminary analysis conducted at PT. XYZ revealed several key issues contributing to this suboptimal performance. The production data indicated that low OEE was primarily caused by high machine downtime, prolonged setup durations, and extended sample collection time. Frequent machine breakdowns occurred due to a lack of preventive maintenance and delayed corrective actions. Setup activities lacked standardization, leading to inconsistencies between operators and frequent delays in production start-up. Similarly, inefficient sample collection processes resulted in bottlenecks that delayed quality verification and disrupted production continuity. These inefficiencies not only reduced equipment utilization but also negatively impacted scheduling, throughput, and overall productivity. In addition, the lack of procedural standardization and poor documentation further contributed to inconsistencies in product quality. This situation highlights the need for a systematic, data-driven approach to process improvement.

To address these challenges, this research adopts the Define, Measure, Analyze, Improve, Control (DMAIC) methodology—an established framework within the Six Sigma approach [3]. DMAIC offers a structured and scientific process for identifying root causes, formulating corrective actions, and sustaining improvements over time [4, 5]. The primary objective of this research is to enhance the operational performance of the injection moulding process at PT. XYZ through the systematic application of DMAIC. The study aims to identify the root causes of low OEE, particularly those related to downtime, setup duration, and sample collection inefficiencies [6-10]. By utilizing analytical tools such as Supplier, Input, Process, Output, and Customer (SIPOC), Fishbone Diagram, Failure Mode and Effects Analysis (FMEA), and the 5-WHYs method, the research seeks to analyze process bottlenecks and implement corrective actions. Furthermore, it evaluates the impact of these interventions on performance indicators, with a focus on increasing OEE to meet or surpass the company's 85% target. Ultimately, this research demonstrates how data-driven continuous improvement methodologies can be effectively applied in real-world industrial environments to improve productivity and process reliability [3, 5, 11].

While DMAIC is a well-established methodology, the novelty of this study lies in its contextual application within an Indonesian injection moulding company that faces resource limitations and high production demands. The findings are expected to provide not only actionable insights for practitioners but also empirical evidence to support the use of structured improvement methodologies in similar manufacturing environments. This paper is organized to present the step-by-step application of DMAIC, supported by operational data and performance metrics, in order to demonstrate the practical value and effectiveness of continuous improvement initiatives in industrial settings.

This study employs a case study approach within the framework of the DMAIC methodology. The case was conducted at PT. XYZ, an Indonesian injection moulding company that manufactures health devices. The purpose of this methodology is to identify process inefficiencies and implement targeted improvements to increase OEE [18, 19]. The implementation of each DMAIC phase is described as follows and is shown in Figure 1.

201.jpg

Figure 1. Research methodology flowchart

3.1 Define phase

The problem definition phase involved preliminary discussions with managers and engineers to identify the core performance issue. The main problem was defined as: “OEE levels consistently below the company’s 85% target, averaging only 81.72%.” To scope the problem, a SIPOC diagram was constructed to map the high-level process flow from suppliers to customers and to clarify inputs, outputs, and key stakeholders.

A Pareto analysis of historical rejection and downtime data was conducted to identify the most critical contributors to performance loss. Interviews and direct observations were used to refine the problem statement into measurable goals for the study.

3.2 Measure phase

In this phase, primary and secondary data were collected to establish baseline performance. Key performance indicators measured included Availability, Performance, Quality, and OEE for three different injection moulding machines (Fanuc, Arburg, Engel), using formulas derived from TPM literature.

Data sources included:

1. Monthly production reports (Jan–Dec),
2. Machine downtime logs,
3. Setup and cleaning records,
4. Sampling and inspection timelines.

Each OEE component was calculated according to the standard formula:

$O E E= Availability \times Performance \times Quality$

The Six Big Losses were categorized and quantified per machine. Data validation was conducted by triangulating logbook entries with supervisor interviews and visual observation of shop floor activities.

3.3 Analyze phase

The collected data were analyzed to identify root causes of low OEE. The Main tools used in this phase included:

1. Fishbone Diagram (to group causes into man, machine, method, and material),
2. 5-WHYs analysis (to drill down into the root of recurring issues),
3. FMEA (to prioritize failure modes using risk priority numbers).

The FMEA was applied to ten critical process steps, with validation sessions conducted through focus group discussions involving supervisors, maintenance teams, and QA personnel. Failure modes with RPN ≥ 48 were selected for intervention.

3.4 Improve phase

Improvement strategies were designed based on the findings from the Analyze phase. Interventions included:

1. Standardization of machine setup procedures,
2. Preventive maintenance scheduling (daily, weekly, and monthly),
3. Implementation of autonomous maintenance,
4. Operator training on drying temperature, mold verification, and product limit checks.

To validate these improvements, pilot tests were carried out over a 3-month period (Oct.–Dec.). Performance metrics before and after improvements were compared to assess effectiveness. Special focus was given to:

1. Changes in OEE across machines,
2. Reduction in idle and setup time,
3. Improvement in yield and reduction in defect rates.

3.5 Control phase

To sustain the improvements, Control Plans were developed, including:

1. Daily check sheets for critical parameters,
2. Weekly review of OEE trends, and
3. Audit checklists to monitor adherence to new standard operating procedures (SOPs).

Visual management tools such as performance dashboards and machine tagging systems were introduced. A Control chart (p-chart) was applied to monitor product quality stability over time. Feedback loops were created through monthly performance review meetings with operations staff.

The detailed flow diagram of the research methodology for this paper is shown in Figure 1.

5.1 Data collection

Before starting the DMAIC stage, it is essential to have a thorough understanding of the injection moulding production process flow being implemented. Figure 3 below illustrates the stages of the injection moulding process.

203.jpg

Figure 3. Process injection molding

The DMAIC approach used is presented in this section.

(1) Define Phase

The define phase starting from identification of the problem and classification of objective.

a) Problem: the OEE rate and performance of moulding company lower than 85%.

b) Objective: The objective of the project was to increase rate of the moulding machine to minimum 85% by eliminating the critical factors resulting loss. The objective to ensures the quality and effectiveness of the company’s products, as well as customer’s satisfaction. In order to identified the process of moulding production, visualization in the form of SIPOC. The SIPOC for this case is presented in Table 2 below.

Table 2. SIPOC

(2) Measure Phase

The OEE is a very established key performance indicator that reflects the performance of a production machine in comprehensive manner. Figure 4 below is the OEE process and formula that illustrates how OEE is calculated, by multiplying the three main components [20].

204.jpg

Figure 4. Modified OEE process and formula

Table 3 provides performance data for a machine, broken down by month, with various metrics that assess machine efficiency and usage. Overall, the machine’s OEE over the entire year is 66.47%, which is moderate but leaves room for improvement. The machine was idle for 206.695 hours, representing significant unused capacity, particularly in months like February, March, and October. A focus on reducing idle time and optimizing production scheduling could improve both run time and OEE in the future. The lowest OEE is in June with the percentage 30.26% and the highest is May 94.50%.

Table 3. Six big losses for injection moulding Fanuc

Table 4 tracks the six big losses that impact the Arburg injection molding machine’s OEE throughout the year. The six big losses include: setup and cleaning time, downtime, idle time, and run time, among other factors. These losses are crucial for identifying inefficiencies and bottlenecks that affect productivity.

1. Idle Time as the Major Contributor to Losses: Throughout the year, idle time was the largest source of inefficiency, with 257.801 hours of idle time. Addressing this could drastically improve productivity.
2. Significant Operational Gaps: The machine was completely idle in January, February, and September, and this contributed to major production losses. Production planning and scheduling should be re-evaluated to avoid long periods of inactivity.
3. Opportunities to Reduce Setup and Cleaning Time: With 4.100 hours spent on setup and cleaning, there is room for optimizing these processes to maximize run time.
4. Strong Efficiency in Productive Months: July and December demonstrated high OEE and efficient machine use. These months could serve as benchmarks for improving other months where OEE and run time were lower.

Table 4. Six big losses mesin injection moulding Arburg

Table 5 outlines the six big losses for the Eagle machine, detailing run time, setup/cleaning time, downtime, idle time, and OEE for each month of the year. From January to June, the machine was completely idle, with 0 hours of run time and an OEE of 0%, leading to substantial idle time. Starting in July, the machine saw significant operation, with a run time of 21.940 hours and an OEE of 42.30%, though downtime was still high at 6.240 hours.

Table 5. Six big losses mesin injection moulding Engel

The run time peaked in December with 32.120 hours and the highest OEE of the year at 87.63%, showing optimal machine usage. Throughout the year, setup and cleaning time remained moderate, with a total of 3.750 hours, while downtime accumulated to 22.760 hours. Idle time, the largest contributor to losses, reached a total of 222.545 hours. The overall OEE for the year was 59.61%, showing room for improvement in machine efficiency and reduction of idle periods.

From the Table 6, one of the potential failures found is the use of the wrong material, which can cause delays or stoppages in production. The main cause of this problem is the lack of identification on the raw material container, with a severity level (S) of 6, a probability of occurrence (O) of 2, and a detection (D) of 4, resulting in an RPN value of 48. To reduce this risk, a control has been implemented in the form of providing identification labels on the material container.

Table 6. RPN calculation

In addition, there is a failure in the form of a difference in color in the product that can be caused by contamination from the supplier. Although it has the same severity level (S = 6), the probability of occurrence (O = 1) is lower, and the detection is quite good (D = 4), resulting in an RPN of 23. The preventive efforts made are to use materials in a batch system to avoid contamination.

Another failure analyzed is the preparation of the wrong mold or tool, which can cause production delays. This is generally caused by the technician's negligence in collecting the appropriate mold, with an RPN of 48. To overcome this, the mold identification and verification steps are implemented with the mold folder before use.

From this analysis, it can be concluded that risks with high RPN (≥48) such as the use of wrong materials and wrong mold preparation should be the top priority in process improvement. Improvement efforts can be made by ensuring a better identification system, improving technician training, and tightening raw material control. Meanwhile, risks with lower RPN, such as product color differences, must still be monitored even though they are not the top priority. With this approach, the effectiveness of the production process can be improved, and potential failures can be minimized.

(3) Analysis Phase

In this phase, the data collected has been analyzed using cause and effect diagram to identify major defects and their causes for addressing them in order to improve the process. Here’s a Fishbone (Ishikawa) diagram breakdown for the problem, highlighting the key causes in different categories as shown in Figure 5.

205.jpg

Figure 5. Fishbone Diagram

1. Man (People):

1. Lack of proper training for operators, leading to inefficiencies in production.

2. Material:

1. Poor quality of raw materials affecting product consistency.
2. Slug weight variation causing defects in the final product.

3. Machine:

1. Improper maintenance of machinery, leading to frequent breakdowns.
2. Mold cleaning not done regularly, impacting product quality and cycle time.

4. Method:

1. High speed of injection, causing defects due to insufficient control.
2. No proper cycle time established, resulting in inconsistent product output.

These issues contribute to inefficiencies and quality problems in the production process.

(4) Improve Phase

In this phase, the primary objective is to develop, test, and implement solutions to address the root causes identified during the Analyze phase. In this phase, the main objective is to develop, test, and implement solutions to address the root causes identified during the Analysis phase. Recommendations for addressing the issues are presented in Table 7.

Table 7. Improvement recommendations

5.2 Control

The objective of the control phase is to ensure the sustainability of the improved and modified system, making it more resilient and well-maintained to keep the process stable. It has been observed that without the implementation of restrictions, the system may become uncontrollable.

This study aims to address the issue of suboptimal performance in injection molding operations at PT. XYZ by applying the DMAIC methodology. The primary objective is to improve OEE by identifying and reducing the root causes of performance loss—especially those related to downtime, long setup times, and sampling inefficiencies.

Through a structured application of DMAIC, supported by various tools such as SIPOC, FMEA, Fishbone Diagram, and 5-WHYs, this study successfully identified the most critical failure modes affecting productivity. Targeted interventions, including the implementation of preventive maintenance, standardization of setup procedures, and enhanced operator training, resulted in a substantial increase in OEE to 87.63%, exceeding the company’s performance targets.

These findings have significant practical implications for manufacturing companies operating in similar environments. This study demonstrates that a data-driven continuous improvement framework, when applied systematically and supported by cross-functional collaboration, can produce measurable performance improvements even in resource-constrained situations.

However, this study also encountered certain limitations. The scope of implementation was limited to a single plant and focused on three machines. Factors such as organizational culture, supplier variability, and human resource constraints were not explored in depth. These aspects may affect the scalability of the improvement strategy.

For future research, it is recommended to:

Expand the application of DMAIC across multiple facilities or product lines to test scalability and consistency, integrate advanced data analytics or IoT-based monitoring for real-time performance tracking, Investigate the long-term sustainability of various control measures through longitudinal studies, and explore the role of human factors, such as the effectiveness of training, motivation, and interdepartmental communication, in the success of operational improvements.

In conclusion, this study contributes to academic understanding and industry practice by demonstrating how a structured methodology such as DMAIC can deliver practical improvements in manufacturing performance. It also provides a replicable framework for other organizations seeking to improve efficiency and quality in complex production environments.