Publication Search

Low productivity and high sick leave in the weaving division of PT XYZ in Yogyakarta are caused by excessive physical and mental workload on Shuttle Loom Unit 2 operators. The productivity of the weaving division of PT XYZ was recorded at only 61.96% (target 75%) with sick leave of 4.17%, indicating operator fatigue. This study measured physical workload using the Cardiovascular Load (CVL) method based on heart rate and mental workload using NASA-TLX, and analyzed the correlation with age and length of service. The results showed a moderate physical workload category (40% light CVL <30, 60% moderate 30-60) at a temperature of 30.5°C, and high-very high mental workload (45% high 50-79, 55% very high ≥80, average EF 79.3) due to strict quality targets (0.5% defects). There is a correlation between CVL and NASA-TLX with age. Recommendations include reducing the daily production target from 100 to 85-88 yards, optimizing ergonomics, and training to reduce EF to 65 and defects to 0.3, to increase effective productivity.

Fatigue is a critical issue in labour-intensive small industries, especially in traditional food production such as tofu manufacturing. This study aims to develop a fatigue classification model using a decision tree algorithm by integrating subjective assessments of the work system through the Macroergonomic Organizational Questionnaire Survey (MOQS) and objective physiological indicators, specifically Cardiovascular Load (CVL). The research was conducted in a tofu home industry located in Kalisari Village, Banyumas, Indonesia. Primary data were collected from 10 workers through MOQS questionnaires and heart rate measurements taken at rest and during work. CVL values were calculated and used as labels for classification into three categories: low, moderate, and high fatigue. Meanwhile, MOQS dimension scores (organization, job, personal, environment, and technology) were transformed into interval data and used as classification features. A decision tree model was built using the CART algorithm and visualized for interpretability. The results show that all workers experienced at least moderate fatigue, with 20% categorized as high fatigue. The decision tree revealed that the dimensions of organizational and personal factors were the most influential in predicting fatigue levels. The model provides a practical and interpretable tool to support decision-making in scheduling, workload balancing, and ergonomic interventions. This study demonstrates a novel approach to combining macroergonomic assessments and physiological data with machine learning for practical fatigue risk management in small-scale food production environments.

A manufacturing company engaged in the furniture industry, specializing in wood, rattan and glass products. Most of the production orders were related to table products, stands and chests, with a total of 15,224 orders for other products. Based on results of direct observation at PT. Romi Violeta from the production process department says that due to the limited working hours, especially as the production orders for tables, shelves, etc. increase every month, the physical and mental strain employees take to meet their production targets. I often complain about. For commodity products. The purpose of the research is to find cases of high physical and mental stress at manufacturing sites with high stress standards such as tables, shelves, and chests of drawers, and to provide suggestions for improving the reduction of physical and mental stress. . Based on an analysis of the results, the mean cardiovascular load (CVL) percentage was found to be 43.902%, indicating that the assessment needs improvement, and the Bourdon Vielsma test speed result was 12.593 seconds with a score of 7. , a body weight score (WS) of 11.5 is included in the adequate group (C). Accuracy is 3.26, score is 8, fair group (CB) weighted score (WS) is 12, constant is 8.185 seconds, score is 5.5, score is . 8 are in the suspicious (R) group.