This study aimed to develop a regression-based model for predicting tool life in manufacturing environments, with goals of enhancing productivity and reducing costs. In machining operations, particularly roughing processes, high cutting forces can accelerate tool wear, often leading to process interruptions and increased defect rates. Previous research on tool life prediction has frequently relied on empirical models and statistical methods, which face limitations in reliability across diverse machining conditions. To address this issue, we proposed a data-driven approach that could collects tool wear data under varying machining conditions (such as cutting speed, feed rate, and depth of cut) and applied regression models to predict tool life effectively. The model’s performance was validated under multiple conditions to assess its predictive accuracy. This study offers a practical tool life management solution for manufacturing settings, optimizing tool usage and enhancing operational efficiency.

In this study, the deformation of a large industrial door subjected to wind load was investigated through computational fluid dynamic and structural analyses. The model for the structural analysis was simplified by considering the PVC curtain and wind bar in the shape of the actual door. The pressure distribution acting on the front of the door was obtained from computational fluid dynamic analysis and the deformation of the door was obtained from structural analysis. According to the results, the pressure distribution was not uniform on the front of the door and varied depending on the location. The distribution of the deflection in the wind bar was obtained and it was found that the position of the maximum deformation occurred slightly above the center of the door. Finally, the deformation of the door could be predicted by analyzing the deflections of the wind bar subjected to different wind speeds through regression analysis.

In the printed circuit board (PCB) manufacturing industry, the yield is an important management factor as it significantly affects the product cost and quality. However, in real situations, it is difficult to ensure a high yield in a manufacturing process, because the products are manufactured through numerous nanoscale manufacturing operations. Thus, for improving the yield, it is necessary to analyze the key process parameters and equipment parameters that result in a low yield. In this study, critical equipment parameters that affect the yield were extracted through a mutual analysis of the equipment parameters (x) and process parameters (y) in the plastic ball grid array (PBGA) manufacturing process. To this end, the study uses the correlation coefficient to apply the heuristic algorithm that extracts critical parameters that keep the redundancy among the equipment parameters to a minimum and exert maximum impact on the critical process parameters. Additionally, by using the general regression neural network technique, the effects of the critical equipment parameters on the process parameters were confirmed. The test results were applied to the PBGA production line and an improvement in the yield was confirmed.

The objective of this study is to investigate a novel temperature and humidity prediction algorithm for smart greenhouse based on the machine learning method. The smart greenhouse is known to increase farm production by automatically controlling temperature and humidity and other factors. However, maintaining constant inside temperature and humidity in the conventional smart greenhouse system is still a problem because of the multiple time delay elements. To solve the problems, prediction control scheme is required. But, since the system is highly nonlinear with the lack of sensory data, predicting accurate temperature and humidity is very challenging. In this paper, the multi-dimensional Long Short-Term Memory networks (LSTMs) is being applied to deal with the unstructured greenhouse environmental data. The designed LSTMs learning model is trained with the 27 dimensional data which comprises of all the greenhouse control parameter and environmental sensory data. The prediction performance was evaluated using the short, mid and long term experiments. Also, the comparison with the conventional recurrent neural networks (RNNs) based prediction algorithm was done using the experimental results and later on discussions.