Skip to main content

Application of neural networks to integrated ergonomics

Article Category

Article available in the folowing languages:

Human movement simulation in industrial environments

Ergonomics, i.e. the study of the relationship between workers and their working environment, proves to be an essential factor in the effort to remain competitive. The current project using elaborate software simulates human motion during an industrial process. The powerful computing toolset developed is currently applicable to genuine industrial problems.

Industrial Technologies

Ergonomics play a major role in increasing the productivity of a given workplace, although, testing the physical ergonomics of systems is costly, time consuming and the methodology applied to one system is hardly ever transferable to a different one. Failure, on the other hand, to perform an ergonomic analysis carries the danger of manufacturing a product that may not be competitive. To aid modern industrial ergonomics a realistic simulation of human movement is needed. To simulate human motion, traditional bio-inspired strategies were usually employed, which cannot take into account the subjective perception of the contextual quantities, an important factor in an industrial environment. In contrast, the current project, in order to perform the simulation, has used data from a collection of physical experiments that are able to incorporate these additional degrees of freedom of human movements. Along with motion analysers, artificial neural networks and open Computer Aided Design (CAD) environments, both real and virtual experiments can be performed in order to assess specific industrial manufacturing scenarios. The CAD representation allows the visualisation of the analysed environment since CAD systems allow an engineer to view a design from any angle, with merely the push of a button. Neural networks on the other hand, a well-known branch of artificial intelligence, are a network of processors designed to mimic the transmission of signals within the human brain. Until now neural networks have been applied successfully in diverse areas like financial market analyses or voice recognition. The incorporation of neural networks in ergonomics is an innovative accomplishment of this project. Application of the developed powerful computer system has shown that human movement data analysis aids considerably the ergonomics testing and is capable of providing a common methodology for such experiments. It is this common methodology, made possible through the use of neural networks, which may eventually bring industrial research and development in this area to the highest level.

Discover other articles in the same domain of application