CONVOLUTIONAL NEURAL NETWORK UNTUK KLASIFIKASI RISIKO POSTUR KERJA PADA PENANGANAN MATERIAL MANUAL
JASMINE MAE LESMANA, Ir. Achmad Pratama Rifai S.T., M.Eng., Ph.D
2024 | Skripsi | TEKNIK INDUSTRI
With the advancement of industrial technology, the use of robotics and automation has expanded across various sectors. However, many tasks, such as manual material handling (MMH), still rely on human labor, particularly in developing countries where full automation is not yet feasible. These manual handling tasks expose workers to the risk of musculoskeletal disorders (MSDs), especially if non-ergonomic postures are adopted over prolonged periods. Traditional ergonomic assessment methods, such as RULA, require manual evaluations that are often subjective and inefficient.
This study aims to develop a deep learning-based method using Convolutional Neural Networks (CNN) to automatically classify ergonomic risk levels associated with MMH tasks. By utilizing image data and RULA score calculations, this approach seeks to design and develop a CNN model to train, test, and classify the safety risk levels of work postures in manual material handling tasks related to musculoskeletal disorders (MSDs). The study uses a static image dataset representing various postures and applies a CNN model to classify risk into four categories.
The results of the study show that the CNN model, particularly MobileNet-V2, achieved an accuracy of 82% and provided consistent predictions across various risk categories. This research contributes to the field of ergonomics by demonstrating how machine learning can enhance safety and productivity in industrial environments.
Kata Kunci : manual material handling, gangguan muskuloskeletal, convolutional neural network, deep learning, klasifikasi risiko postur kerja