I declare that this project report titled "DISCRETE EVENT SIMULATION AS A DECISION-MAKING TOOL FOR IMPROVING OVERALL LINE EFFICIENCY" was prepared by YONG WEN YAO and has met the required standard for submission in partial fulfillment of the requirements for the award of Bachelor of Engineering (Honours) Industrial Engineering from Universiti Tunku Abdul Rahman. 62 Table 4.7: Simulation results for printer configuration steps 64 Table 4.8: Simulation result of corrugated board model 65 Table 4.9: Attribute values and difficulty of each.
Introduction
Background
Decision Making
From a technical perspective, decision-making involves a more holistic approach, rather than relying on intuition or personal experience. For example, Sudhagar (2017) used Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), a multi-criteria decision-making tool, to identify the best alternative solution to improve one of the processes in aluminum alloy production.
Discrete Event Simulation
Schruben (2000) introduced that the dynamics of a discrete event system can be expressed through mathematical programming formulation to derive an optimal solution to the optimization problem. This is because most of the complex production systems can be modeled using software and the model development time is also reduced.
Overall Equipment Effectiveness
Consequently, improvements can be made to address the detected losses in order to improve system performance. This is because level of improvement can be evaluated by comparing future OEE value with initial OEE value to justify the effectiveness of improvement applied (Dal et al., 2000).
Problem Statement
In this case, projects such as production planning and decision-making for process improvement must be carried out remotely. Therefore, the adoption of software in decision making for process improvement in the manufacturing system should be promoted.
Aim and Objectives
Scope
Therefore, using Minitab to analyze statistical data is fast and in addition, data in Excel sheet can be easily imported into Minitab. Last but not least, a box manufacturing factory is chosen as the case study's background because its manufacturing system is considered to be moderately complex due to workshop manufacturing.
Thesis Outline
It is an advanced and intuitive data analysis tool, as data can be easily entered, manipulated and thoroughly interpreted in various forms of statistical analysis, graphs and charts. The nature of shop floor manufacturing has created many optimization problems in planning and material handling, as WIP paths vary between product families.
Introduction
Review of Decision Making
In addition, Papakostas (2012) had developed an agent-based methodology to address decision-making problems in the production system. There were 8 agents in total and global decision making was done on the 8th agent.
Review of Process Improvement
Therefore, TOPSIS and GRA were used to determine the optimal values of the process parameters to produce high-quality products. An OLE enhancement scheme was designed that adopted a Fourier amplitude sensitivity test to determine the essential parameters to focus on.
Review of Decision Making for Process Improvement
A systematic methodology synthesizing productivity optimization, quality control, and cost minimization was found to be lacking. Through the literature review, it was found that computer simulation is a popular approach used to design from scratch, test or modify lean system.
Analysis of Literature Reviewed
Discussion of Literature Reviewed
However, there is a lack of a structured framework that embodies considerations of process improvement in strategic decision-making frameworks. However, these factors are largely absent in all the frameworks reviewed regarding decision-making for process improvements.
Introduction
Flow of Research
However, key elements of process improvement, such as performance benchmarking and root cause analysis, must also be incorporated into the decision-making basis to find the best alternative that can effectively improve the performance of a production system. Subsequently, with respect to the second objective, a comprehensive decision framework for process improvements was formulated by synthesizing the strengths and addressing the gaps in the reviewed frameworks. Finally, a conclusion is drawn on the feasibility of the proposed decision-making tool and framework to improve the performance of a production line.
Development of Framework
A case study was conducted by applying the proposed framework in a real production system for validation. To carry out the case study, data was collected on site and a simulation model of the production system was constructed using WITNESS software with reference to the collected data as input values. Suggested frameworks were followed to seek the best implementation that optimizes the production system performance measures.
Stage 1 - Initiation
It is important to ensure that the knowledge about the manufacturing system is sufficient to carry out further investigation. The former refers to symptoms or facts that reflect the poor performance of the manufacturing system. The fourth step is to define process improvement objective according to the problems or improvement opportunities observed in the manufacturing system and performance measure.
Stage 2 - Modelling
The third step is to build a simulation model of the production system using DES software. The simulation model is then run and the necessary performance data is collected for the next step. Furthermore, if the model verification or validation fails, it means that the simulation model is inaccurate.
Stage 3 - Analysis
In addition, methodologies that can be employed to analyze root causes are Ishikawa diagram, 5 Whys and 4 Ms principles. Pareto analysis can be used to identify the most critical root cause that has the greatest impact on the production system and prioritize that root cause for improvement. By focusing on the main cause, the result of an improvement project can be maximized (Cheah et al., 2020).
Stage 4 - Selection
This is to observe, compare and analyze the main effects of each solution on the performance measures of the production system, as well as the interaction between the attribute values of the different solutions. This step is to establish consistency in the values of the performance measures between the decision criteria so that they can be compared with each other. While equation 3.7 shows the method to normalize the performance measures in the indirect category, in which the system performance drops as the measure increases.
Stage 5 - Termination
The fourth and final step in this phase is to calculate the overall performance score for the existing system as well as each proposed solution, and select the alternative with the best performance. To calculate the overall performance score, the performance of each decision criterion is multiplied by the weight of the corresponding criterion and the individual decision criterion scores are summed as shown in equation 3.8. However, if the experimental result of all alternatives is unsatisfactory, in which the overall performance results of all alternatives are lower than those of the existing system, Phase 3 should be repeated to analyze the problem again to search for effective solutions .
Introduction
Background of the Company for Case Study
This means that although the product variety is the same, the product design may be different for different customer orders in terms of size, print design, paper stock, etc. Ultimately, each product variety has a unique production route through the factory and each product design requires distinctive materials and tools during production.
Stage 1 - Initiation
According to row (i) = 1 and 2 ≤ column (j) ≤ 5 in Table 4.7(a), the values of p1j were greater than 1, as this indicates that OLE is more important than other decision criteria because the main objective of this project . Based on Table 3.2, SA is equal to 1.12 because there were 5 decision criteria in this project. A slightly important change in the work process. The result is likely to deviate slightly from the simulated results due to uncontrolled aspects.
Stage 2 - Modelling
Next, a simulation model of the production line was constructed based on process flow illustrated in Figure 4.1 and information compiled in Figure 4.2. Averages of the performance measures were recorded as shown in Table 4.5 by running 5 replications of the simulation model. As shown in Figure 4.6(b), the theoretical time for the last part sent from the last workstation corresponded to the end time of the deterministic model, which was 727,545. minute.
Stage 3 - Analysis
The new values of total setup time and average setup waste will be fed into the line simulation model to obtain the overall performance of the virtual production line. The variables to change for each run (as shown in Table 4.9) were added to the simulation model because they were input parameters of the experimenter function. Based on the overall performance scores calculated for all combinations of alternatives, it was concluded that increasing the efficiency of setting up printers and improving the problem-solving ability of operators were the most desirable implementations because it had the highest score of overall performance, which was 0.9219 as shown in table 4.12.
Stage 5 - Termination
In order to improve the problem-solving skills, training sessions should be held for the operators to cultivate their skills, knowledge and ingenuity in solving machine problems that affect product quality. The HR department should organize periodic training sessions for the operators, and production personnel experienced in handling machine problems should devise strategies for improving the operators. To achieve the target MTTS value and overcome the problems associated with LOD, production supervisors must constantly monitor operators.
Findings and Discussion
As a result, the superiority of the solutions is tangible and the decision making is done based on the overall performance scores of the alternatives. Apart from that, comparing the conventional way with the new framework, the new framework emphasizes the quality of the solution chosen for implementation. Furthermore, the new framework differs in that it facilitates improvement projects involving different aspects or workstations and predicts the long-term effects of the solutions on the production line.
Introduction
Summary of Research
As a result, important decision-making elements as well as process improvement frameworks were identified. If multi-criteria decision making (MCDM) is involved, the analytic hierarchy process (AHP) pairwise comparison method will be used to assign criterion weights to the decision criteria. In this research, a new decision-making framework was formulated after identifying the key features of the decision-making framework in the literature review.
Recommendations for Future Research
All this proved the feasibility of simulation software in decision-making, as it is possible to search for an optimized solution in which goals are achieved with minimal resources. Decision Making in a Manufacturing Environment: Applications of Graph Theory and Fuzzy Multi-Attribute Decision Making Methods. A multi-criteria decision-making approach for process improvement in aluminum alloy friction stir welding.
