Gesture Recognition–Based Smart Training Assistant System for Construction Worker Earplug-Wearing Training

Abstract

Thousands of construction workers suffer noise-induced hearing loss (NIHL) every year from excessive noise exposure on the job, which impairs the quality of their lives and increases the risk of injury. Properly wearing earplugs is very important onsite for worker hearing protection. However, the training provided in the current practice is minimal. Therefore, there is a need to develop an efficient and effective self-training method that can provide both accurate step-by-step earplug-wearing instructions and timely feedback through monitoring. With the development of artificial intelligence and wearable sensor technologies, the possibility of developing an advanced intelligent training method becomes plausible. Therefore, the objective of this paper is to develop a gesture recognition–based smart training assistant system that can automatically evaluate workers’ performance during their earplug-wearing self-training and provide timely feedback to rectify any mistakes. Through the system feasibility test and performance evaluation, the results show that the proposed system can achieve around 90% training accuracy and around 80% testing accuracy recognizing the classified forearm gestures of wearing earplugs for noise protection training using the developed artificial neural network (ANN) models for both hands. The proposed gesture recognition–based smart training assistant system will eventually help industries to improve the performance and safety of employees with low implementation costs.