Development of Artificial Neural Networks Model to Determine Labor Rest Period Based on Environmental Ergonomics
Penulis/Author
Rosa Amalia, S.T.P., M.Sc. (1); Prof. Dr. Mirwan Ushada, S.T.P., M.App.Life.Sc. (2); Dr. Agung Putra Pamungkas, S.T.P., M.Agr. (3)
Tanggal/Date
28 2023
Kata Kunci/Keyword
Abstrak/Abstract
Food SMEs (Small and Medium Enterprises) was an example of labor intensive industry, which involved many labors to pursue the production activities. Food SMEs required complex process in production activities. A support to increase the work productivity and reduce ergonomic risks of the activities was required. The study was conducted at Tofu SMEs. Determination of rest period could be developed to give some recovery times to labors. WBGT (Wet Bulb Globe Temperature) was estimated to determine the amount of rest period. Rest period was determined by workstation environment and workload labor. ANN (Artificial Neural Networks) model was carried out due to a non-linear relationship. ANN was used to store the information from data set and predict the amount of rest period and WBGT. ANN was employed using backpropagation. Backpropagation algorithm used the error value to change the weight with forward propagation then backward propagation. The result showed that dry bulb temperature, heart rate, wet bulb temperature, and gender had significant impact toward rest period and WBGT. A 180 data set from tofu SMEs was divided into training data (80%) and validation data (20%). The optimal ANN structure was determined of 4 input, 4 hidden, and 2 output neuron. The activation function was sigmoid for both layers. SSE (Sum of Squared Error) was used to obtain the best structure. The value of R2 was equal to above 0,900 which indicated that ANN could be used to model labor rest period based on environmental ergonomics.
