Wardhani, Anggita Kusuma (2026) Analisis Beban Kerja Fisik Berdasarkan Suhu Permukaan Kulit Berbasis ROI dan Cardiovascular Load pada Aktivitas Sepeda Statis. Other thesis, Institut Teknologi Sepuluh Nopember.
|
Text
5010221040-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (5MB) | Request a copy |
Abstract
Hampir seluruh sektor pekerjaan di Indonesia berhubungan erat dengan aktivitas fisik dengan tingkat beban kerja yang berbeda-beda. Sehingga perlu adanya pengukuran beban kerja fisik agar dapat dilakukan penyesuaian beban kerja serta mengurangi risiko kecelakaan kerja yang merugikan bagi pekerja maupun perusahaan. Metode pengukuran beban kerja fisik yang umum digunakan adalah Persentase Cardiovascular Load (%CVL) karena berbasis denyut jantung. Metode ini memerlukan kontak langsung dengan tubuh karena menggunakan wearable sensor untuk merekam denyut jantung. Selain berpotensi mengurangi kenyamanan pengguna, pelepasan perangkat selama proses pengukuran dapat menyebabkan kontinuitas pemantauan data terganggu. Oleh karena itu, diperlukan metode alternatif pengukuran non kontak, salah satunya adalah Thermal Imaging yang menangkap suhu pada tubuh manusia dengan menggambarkan peningkatan atau penurunan suhu tubuh yang selaras dengan peningkatan atau penurunan denyut jantung. Penelitian ini bertujuan untuk menganalisa kemampuan suhu permukaan kulit pada beberapa Region of Interest (ROI) untuk mengklasifikasikan tingkat beban kerja fisik pada aktivitas sepeda statis terhadap hasil klasifikasi beban kerja fisik %CVL. Penelitian ini menggunakan desain eksperimen dengan tiga skenario yang dilakukan oleh 29 responden laki-laki sehat yang melakukan aktivitas sepeda statis dengan intensitas beban yang meningkat. Beban yang diatur setara intensitas ringan atau setara dengan 3.5 MET, sedang atau setara dengan 6.8 MET (Metabolic Equivalent of Task), dan berat atau setara dengan 8.8 MET serta suhu ruangan yang dikontrol pada suhu 27-28oC. Suhu tubuh ditangkap menggunakan kamera termal FLIR dan denyut jantung direkam menggunakan chest strap Polar. Pengujian diawali dengan uji normalitas data, lalu uji korelasi, data training dan testing untuk melakukan klasifikasi tingkat beban kerja fisik. Hasil penelitian menunjukkan bahwa suhu pada seluruh ROI berbeda secara signifikan antar tingkat beban kerja karena nilai p-value<0.005. Suhu permukaan kulit mengalami penurunan seiring peningkatan intensitas beban karena termoregulasi. Hasil klasifikasi dan uji akurasi, presisi, recall, dan f1 score menunjukkan bahwa ROI lumbar merupakan ROI tunggal yang paling representatif, sedangkan kombinasi ROI scapula, lumbar, dan dada merupakan ROI gabungan yang paling representatif dalam mengklasifikasikan tingkat beban kerja fisik.
================================================================================================================================
Nearly all occupational sectors in Indonesia are closely linked to physical activity, with varying levels of physical workload. Therefore, it is necessary to measure physical workload in order to adjust workloads and reduce the risk of workplace accidents that are detrimental to both workers and companies. A commonly used method for measuring physical workload is the Cardiovascular Load Percentage (%CVL) because it is based on heart rate. This method requires direct contact with the body, as it uses wearable sensors to record heart rate. In addition to potentially reducing user comfort, removing the device during the measurement process can disrupt the continuity of data monitoring. Therefore, an alternative non-contact measurement method is needed; one such method is Thermal Imaging, which captures body temperature by depicting increases or decreases in body temperature that correspond to increases or decreases in heart rate. This study aims to analyze the ability of skin surface temperature in several Regions of Interest (ROIs) to classify the level of physical workload during stationary cycling activities against the physical workload classification results in %CVL. This study employed an experimental design with three scenarios involving 29 healthy male participants who performed stationary cycling activities at increasing intensity levels. The load was set to light intensity, equivalent to 3.5 METs, moderate. The workload was set to light intensity (equivalent to 3.5 METs), moderate intensity (equivalent to 6.8 METs), and heavy intensity (equivalent to 8.8 METs), with the room temperature controlled at 27–28°C. Body temperature was captured using a FLIR thermal camera, and heart rate was recorded using a Polar chest strap. The study began with a normality test of the data, followed by a correlation test, and the use of training and testing datasets to classify physical workload levels. The results showed that temperatures across all regions of interest (ROIs) differed significantly across workload levels, as indicated by p-values < 0.005. Skin surface temperature decreased as workload intensity increased due to thermoregulation. The results of the classification and tests for accuracy, precision, recall, and F1 score showed that the lumbar ROI was the most representative single ROI, while the combination of the scapular, lumbar, and chest ROIs was the most representative set of ROIs for classifying physical workload levels.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Beban Kerja Fisik, Thermal Imaging, Cardiovascular Load (%CVL), Denyut Jantung, Kelelahan Kerja Fisik, Physical Workload, Thermal Imaging, Cardiovascular Load (%CVL), Heart Rate, Physical Fatigue |
| Subjects: | Q Science > QP Physiology > QP34.5 Human engineering T Technology > T Technology (General) T Technology > T Technology (General) > T55.3.H3 Hazardous substances--Safety measures. |
| Divisions: | Faculty of Industrial Technology and Systems Engineering (INDSYS) > Industrial Engineering > 26201-(S1) Undergraduate Thesis |
| Depositing User: | Anggita Kusuma Wardhani |
| Date Deposited: | 31 Jul 2026 03:57 |
| Last Modified: | 31 Jul 2026 03:57 |
| URI: | http://repository.its.ac.id/id/eprint/140676 |
Actions (login required)
![]() |
View Item |
