USE OF FACIAL LANDMARKS MOVEMENTS DETECTION AS A MENTAL WORKLOAD MONITORING SYSTEM IN THE TASK OF AIR TRAFFIC CONTROL

• Main Author: ABDUL AZIS, HASTIAN
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/39822
• Institution: Institut Teknologi Bandung

Air Traffic Control (ATC) is a job that has a high level of responsibility and risk. Amount of traffic congestion that occurs in Indonesia has increased the mental workload of ATC in controling the traffic as well. Considering problems regarding the higher mental workload, ATC needs a real-time workload monitoring system that can detect the workload conditions experienced by the workers. Before the experiment, there was the preliminary survey phase. Each participants at preliminary surver was required to fill a questionnaire about the mental workload carried out by ATC officers in Jakarta Air Traffic Service Center (JATSC) Soekarno-Hatta International Airport. This research, carried out by laboratory experiment, was conducted to improve the use of facial landmarks movements detection as a mental workload monitoring system in ATC. These experiments used the ATC simulator with 12 male participants between 19-23 years of age with undergraduate education in engineering. A measured called the National Aeronautical and Space Administration Task Load Index (NASA-TLX) was used to determine the level of mental workload that contained subjective opinions. The experiment was carried out for three hours in each task demands. There are two kinds of task demands, task with high amounts of traffic and tasks with a medium amounts of traffic. NASA-TLX data collection was carried out at the end of each task completion at the predetermined level of difficulty. The main data in this study is the movements of facial landmarks. Processing of facial landmarks movements data is developed based on data recording of facial activities of participants for two minutes at each of a four-time assignments. The assignments are the initial experiment (0-2 minutes), the first hour (58-60 minutes), the second hour (118-120 minutes), and the third hour (178-180 minutes). Data obtained from the facial landmarks movements detection software is further processed to obtain data on movement averages and a number of changes direction move (for x-axis and y-axis). In this study 23 landmarks were used on the jaw, eyebrows, eyes, nose, and mouth. The result of this study reveals that the characteristics of ATX officers’ mental workload at JATSC were not significantly affected by work shift and work units. Detection of landmark face movements can distinguish working conditions in high and medium traffic conditions. Moreover, sensitivity tests revealed that eyebrows and eyes facial landmarks are the indicators that can serve as a sensitive indicator for distinguishing mental workloads.