Using Applied Computing on Embedded Computers to Build Digital Twins in a Fog Computing Environment

Abstract

This article describes the construction of digital twins based on flexible threads using applied computing on embedded (technological) computers. An overview of computing on embedded computers is presented. The problem is identified as the need to obtain additional information from sensors for building digital twins. The place of sensory information in event diagnostics is determined. The approach of information collection policies was applied to build digital twins that are described on several levels with detailing on each level. A digital twin of the train and each subway car was built using a template of a typical subway passenger car and a locomotive car. Several levels of digital twins' presentation have been proven effective. The use of fog computing made it possible to increase the speed of building a digital twin of the first level by 3.17 times, the detail representation of subsequent levels by 3.91 times, and the accuracy of determining events by 11.2%. This method can be applied not only to subway cars, but also to suburban trains, high-speed peregrine falcons, allegro, swallows, city trams and other public transport.