EVALUATION OF FATIGUE LEVEL BY SAFE ENTRY STATION USING NOVEL DEEP LEARNING TECHNIQUE

Abstract

Fatigue is an important component for screening for “Fit for Duty” at work place. The main objective of this research paper is to identify a novel deep learning technique that can be used to screen fatigue in workplace setting. In order to achieve personal and professional goals, enhance the structure of the organization and to sustain one’s living conditions in an appropriate manner, it is necessary to take into consideration the aspects of health and safety of the employees. This research paper outlines our work on the importance of detecting health, safety and fatigue in the workplace with state-of-art technology before even starting a job.
 This paper proposes a real-time comprehensive employee fatigue detection algorithm based on different facial landmarks to improve the detection accuracy, which detects the employee’s fatigue status by using facial video sequences without equipping them with sensor devices. The facial area is analyzed including detection of left and right eye along with the mouth region.
 In this paper we are proposing a novel deep learning technique to classify high, mid and low levels of fatigue. We are performing this activity at a safe entry station (SES) which also measures other vital parameters such as Body Temperature, Eye Redness, Heart Rate and Respiration Rate. The focus of the current study is on fatigue detection and our AI pipeline achieved 91% accuracy on data points collected at various sites in identification of fatigue levels.