International Journal of Quality Engineering and Technology (5 papers in press)

Regular Issues

• Harnessing deep learning for quality engineering and technology: innovations in process optimisation, defect detection, and predictive quality control  
  by Pratik Patel, Swagata Sarkar, N. Ashok Kumar, Tanvi Jaydeep Patel 
  Abstract: If you are in charge of water sources, you need to be able to guess how streams will flow. We can learn a lot from this study about how well complicated deep learning models can guess when the Gilgit River Basin’s water level will be high and low every month. CNN-LSTM, CNN-BiLSTM, CNN-GRU, CNN-BiGRU, LSTM, BiLSTM, and GRU were all employed. Each of the final four is a combination of these. The model did well for our study based on its RMSE, MAE, NSE, and R2 marks. There’s a problem. R2 tells you how strong a link is. Simple models like LSTM and GRU did not do as well with that data. But the mix models did a lot better. CNN-BiGRU and CNN-BiLSTM did the best most of the time. It was taught with an R2 of 0.962 and tested with an R2 of 0.929. It got 144.1%, which was good enough for second place. CNN can help you find things in space. Now, things have a better chance of going well.
  Keywords: long short-term memory; LSTM; gated recurrent unit; GRU; background; CNN-Bi LSTM.
  DOI: 10.1504/IJQET.2025.10070208
   
  
• A robust optimisation model for a dual-objective software reliability growth model with multiple fault types  
  by Mohammadreza Namdar, Rassoul Noorossana 
  Abstract: Reliability and development costs are two significant criteria that highly dependent on software release time and test termination time. Simultaneous optimisation of both criteria is a substantial challenge in current projects. Studies in the literature usually suffer from some of the following drawbacks; 1) considering a single objective function (reliability or development costs); 2) not considering the uncertainty of the parameters; 3) considering one type of software fault, whereas multiple fault types with different debugging costs can occur in real condition; 4) ignoring the discounted rate in cost optimisation. This paper presents a dual-objective robust optimisation method for joint optimisation of the development discounted costs and reliability considering the various fault types with different debugging costs to cover the mentioned weaknesses. Eventually, it analyses the model performance through a case study. The findings prove the distinguished role of uncertainty and the interest rate on both objective functions.
  Keywords: robust optimisation; software reliability; multiple fault types; dual-objective mathematical model; cost optimisation.
  DOI: 10.1504/IJQET.2025.10071173
   
  
• Degradation model selection using a case-based reasoning system  
  by Maryam Kheirandish, Rassoul Noorossana, Mohamad Reza Nayebpour, Hossein Najmizadehbaghini 
  Abstract: The use of degradation data for reliability analyses of products has gained popularity over the past few decades. This approach to reliability analysis reduces the time required to assess the reliability of certain products. However, identifying a suitable model that best fits the degradation data is time-consuming and highly dependent on the practitioner’s experience and the amount of available information in the knowledge database. In this study, an expert system utilising case-based reasoning was developed to reduce the time required for reliability analysis. This system is built on the R5 model, where the retrieval phase employs the K-nearest neighbour algorithm with fuzzy weights to identify similarities. Several studies were reviewed to extract effective features influencing the form of degradation models. The performance of the retrieval phase was simulated to demonstrate its practical usefulness. An initial case base, consisting of real cases, was created for the initial use of experts.
  Keywords: degradation models; case-based reasoning; CBR; K-nearest neighbour; fuzzy weighing; reliability.
  DOI: 10.1504/IJQET.2025.10071174
   
  
• Integrated deep and classical machine learning for ripeness classification of oil-palm fresh fruit bunches  
  by Rindi Kusumawardani, Nani Kurniati, Muhammad Kalif Qisthi Kamil 
  Abstract: The palm oil industry plays a significant role in the global market, driving continuous innovation and improvements in production efficiency. A critical stage in the palm oil production process is the sorting of fresh fruit bunches (FFB) at the loading ramp reception area, which directly impacts product quality. However, the current manual inspection methods rely on single-visual assessments, often leading to inconsistencies and misclassification of FFB ripeness. Inaccurate classification, particularly of unripe FFB, can affect both the yield and quality of crude palm oil (CPO). This study introduces the application of computer vision technology to enhance the FFB sorting process. By leveraging distinct visual characteristics such as colour and the presence of detached fruits, the study integrates deep learning (CNN) and machine learning (HOG-SVM) techniques for improved classification accuracy. The proposed system achieves an overall accuracy of 90% with an average processing time of approximately 30 seconds, demonstrating a significant enhancement over traditional manual sorting methods. This advancement offers a more efficient and reliable approach to FFB inspection, ultimately contributing to improved processing efficiency and product quality.
  Keywords: fresh fruit bunches; FFBs; convolutional neural network; CNN; palm oil; maturity classification; automated inspection; crude palm oil; CPO.
  DOI: 10.1504/IJQET.2025.10071695
   
  
• Reliability comparison of the shafts when stress and strength follow exponential and normal distribution subjected to twisting and bending moments  
  by Md. Yakoob Pasha, M. Tirumala Devi, T. Sumathi Uma Maheswari 
  Abstract: Shaft is a rotating machine component that is used to transmit power from one place to another. The design of a shaft is essential, subject to its strength and stress. This paper presents the reliability analysis of the shaft according to: 1) maximum shear stress theory; 2) maximum normal stress theory, for which stress and strength follow exponential and normal distributions. Reliability is computed when changing the twisting moment, bending moment, diameter and strength of the shaft. Reliability comparison has been made when stress and strength follow exponential and normal distribution. The comparison between normal and exponential distributions for designing a shaft under twisting and bending moments offers several advantages. For stress analysis, the normal distribution helps assess shaft reliability under consistent conditions, where the material undergoes gradual bending and twisting over time. In contrast, the exponential distribution identifies reliability in sudden overloads. In terms of design optimisation, comparing these distributions enables engineers to balance shaft design between gradual degradation and sudden overload risks. When predicting lifespan, the normal distribution provides a broader understanding of long-term behaviour under sustained stress, while the exponential distribution focuses on early life or sudden failures. Using both approaches together supports more optimised and comprehensive designs.
  Keywords: reliability comparison; twisting moment; bending moment; combined twisting and bending moment; exponential distribution: normal distribution: maximum shear stress theory: maximum normal stress theory: round solid shaft.
  DOI: 10.1504/IJQET.2025.10071696