Implementation of Digital Twinning of Multiple Sensors on a Cloudlet System built using Qualcomm Snapdragon 410c

Industrial Internet of Things and Industry 4.0 has attracted much attention recently. A significant roadblock in Industrial IoT systems is the data privacy. The data collected from a manufacturing unit is confidential and usually large in volume, with data variations happening at high speeds. The latencies are high when the industrial system is connected to the cloud based on the bandwidth availability. Various paradigms such as Fog, Edge, Dew and Cloudlet systems have emerged over the last few years as an alternative to clouds. Edge computing would require a single powerful processing unit to handle large amounts of data, not only in terms of volume but also in terms of speed. Hence, cloudlets, which are closely related to edge computing but can be built using multiple medium-scale processing units such as System on Chips, provide an attractive alternative to clouds. In this paper, we have built a cloudlet system using multiple Snapdragon 410c and implemented a digital twining system on it. Any form of digital twining requires a powerful graphic processing unit. However, we have been able to achieve the same level of accuracy using a distributed cloudlet framework. Cloudlet system opens various avenues in the usage of IoT for industrial manufacturing systems, starting from design to predictive maintenance and also offers a scalable solution to cater to varying scales of industries, from small scale to industry complexes.