• 09 October 2024
• By: Humas PNM

Prevent Work Accidents - PNM Students Innovate to Create a Deep Learning Based Identification and Monitoring System

Photo: The Derider team and supervisors take a group photo in front of the Robotics Lab

PNM – Madiun State Polytechnic (PNM) students never stop innovating. Passing funding for the 2024 Vocational Student Creativity Program (PKM) Funding Scheme, PNM Computer Control Engineering (TKK) students who are members of the Derider Team developed a sophisticated system based ondeep learningwhich is designed to identify potential dangers and negligent behavior of grinding machine workers in the manufacturing industrial sector.

Based identification and monitoring systemdeep learningThis was developed by the PNM Derider Team which consists of Raisa Zahra Salsabila, Desna Fitria Devi, Faiza Pramudia Ardani, Rama Putra Adithya, Yesica Stefany Yuniar Tanriana. In creating this system, PNM TKK Study Program students were accompanied by supervisor Sulfan Bagus Setyawan, S.ST., M.T.

"This innovation aims to prevent work accidents that often occur due to inappropriate use of personal protective equipment (PPE) and unsafe behavior in the work area." Said Raisa Zahra, head of the Derider Team.

According to data from BPJS Employment, the number of work accidents in the manufacturing industry sector continues to increase, with the largest contribution caused by worker negligence. "In this case, the grinding process is one of the work areas that has a high risk, especially when workers do not use adequate PPE or are in a dangerous position." Raisa explained.

To reduce this risk, the PNM Derider Team innovated to create a System for Identifying and Monitoring Potential Hazards and Negligent Behavior of Grinding Machine Workers Based onDeep LearningEfforts to Prevent Work Accidents. It is hoped that this innovation will improve safety in the workplace.

Photo: Prototype created by Tim Derider

Raisa explained that the system uses YOLOv5s (You Only Look Once) technology, a modeldeep learningwhich was specifically developed for object detectionreal-time. By using cameras, this system monitors workers who are operating grinding machines, detecting whether workers have used PPE such asface shield,masker,earmuff,and gloves. Apart from that, this system is also able to identify worker behavior, such as bending or squatting positions that last for a certain duration, which can increase the risk of injury.

Furthermore, if the system detects negligence such as workers not using PPE correctly or being in an inappropriate working position, a notification will automatically be sent via the Telegram application to the supervisor. In addition, detection results, such as violations of the use of PPE or noise levels exceeding the threshold, will be stored insidedatabaseand can be accessed viawebsitewhich is integrated with the monitoring system.

"Through this innovation, Madiun State Polytechnic students are committed to continuing to contribute to the development of technology that supports work safety in the industrial sector, especially in high-risk areas such as grinding." he concluded. *(Derider/PRIP PNM).

---