Enabling Edge Computing in an IoT-Based Weather Monitoring Application

Abstract

Internet of Things (IoT) applications employ several sensors for gathering data. These sensors are often placed at hard to reach locations since they need to be left unmonitored. Sensors collect data in real time and send it to a cloud server for further processing. Often the sensors generate a large volume of data which is redundant in nature. Transferring sensor data to a cloud server for processing leads to high bandwidth consumption, delay, and increase in operational cost and data security issues. Edge computing allows the sensor data to be stored and analyzed on an edge device, and only the data summary is sent to the cloud server. In this paper, an edge computing approach for managing and analyzing data in a weather monitoring application is proposed. The application has been built using a Raspberry Pi system. The computation has been performed by creating Microsoft Azure IoT Hub resource and Microsoft Azure IoT Edge solution. Results have been compared with AzureML cloud platform.