Automatic Ergonomic Risk Assessment Using a Variational Deep Network Architecture

Abstract

Ergonomic risk assessment is vital for identifying work-related human postures that can be detrimental to the health of a worker. Traditionally, ergonomic risks are reported by human experts through time-consuming and error-prone procedures; however, automatic algorithmic methods have recently started to emerge. To further facilitate the automatic ergonomic risk assessment, this paper proposes a novel variational deep learning architecture to estimate the ergonomic risk of any work-related task by utilizing the Rapid Entire Body Assessment (REBA) framework. The proposed method relies on the processing of RGB images and the extraction of 3D skeletal information that is then fed to a novel deep network for accurate and robust estimation of REBA scores for both individual body parts and the entire body. Through a variational approach, the proposed method processes the skeletal information to construct a descriptive skeletal latent space that can accurately model human postures. Moreover, the proposed method distills knowledge from ground truth ergonomic risk scores and leverages it to further enhance the discrimination ability of the skeletal latent space, leading to improved accuracy. Experiments on two well-known datasets (i.e., University of Washington Indoor Object Manipulation (UW-IOM) and Technische Universität München (TUM) Kitchen) validate the ability of the proposed method to achieve accurate results, overcoming current state-of-the-art methods.

Development of a new ergonomic risks management algorithm on the example of drivers

“…The main disadvantage of all mentioned approaches is the lack of experienced observers, as well as the fact that the obtained results can be subjective [19]. To avoid this, various sensor technologies are implemented, including a system of sensor controllers, which clearly allow determining individual risk [20]. However, the specified approach does not allow to extend the obtained dependencies to work performed in other conditions, as it requires an appropriate check for adequacy [21].…”

“…The analysis of the conducted studies [9][10][11][12][13][14][15][16][17][18][19][20][21] showed that in general there are two main approaches to the ER assessment. The first is based on the defined physical activity of the worker, thanks to the use of different time scales.…”

A comparative neural networks and neuro-fuzzy based REBA methodology in ergonomic risk assessment: An application for service workers

“…The proposed integrated method plays a role as an early warning system by considering risky positions and presenting immediate output to analysts or decision-makers. In addition, few studies integrate image processing and different artificial intelligence methods with ergonomic assessment tools like REBA or RULA (Chatzis et al, 2022;Estrada-Lugo et al, 2022;Paudel and Choi, 2020).…”

“…Artificial neural networks (ANNs) were first proposed by Warren McCulloch and Walter Pitts by following the learning, inference and prediction principles of the human brain (Chatzis et al, 2022). Its structure is inspired by neural networks and presented with mathematical principles for input, output and hidden layers.…”

Mobile Industrial Robotic Vehicles: Navigation With Visual SLAM Methodologies

“…It will monitor the various components needed to complete an order and check if the worker placed that component for boxing. Another spot will be used for monitoring the workers and their ergonomic posture [44]. A report of indicators describing a worker's right and wrong posture will be handed to the management team in order to improve the worker's safety procedures and avoid further health problems.…”