Today, the Internet of Things (IoT) is no longer just a fancy topic. It’s becoming a part of every business, automating and optimising processes in a way impossible in the past. The development of IoT technologies in the cloud enables companies to gather huge volumes of data and use it to monitor and automate infrastructure operations, perform big data analysis or train machine learning algorithms.
HeleCloud professional Radoslav Yovchev not only introduced an overview of the hottest AWS IoT services at the 10th AWS Bulgaria User Group meetup on 12 February 2019 but also demonstrated a short demo of how AWS IoT Core service is integrated with other AWS services such as AWS DynamoDB.
“Each system has its unique requirements which have encouraged the development of a rich diversity of IoT telecommunication technologies”, commented Radoslav. Regardless of the underlying system design, once gathered, the information can be sent to the Cloud by using messaging protocols. AWS mainly operates with protocol MQTT, which is very effective in terms of resources and energy usage.
“The MQTT topics design is crucial when it comes to the manageability and scalability of an IoT system.”, he added. Radoslav recommends designing topics in a way that enables granular categorisation of the IoT things and their functions. Possible topic components include a message type, location, device type, device model, etc.
“Attach the policy to groups and not to the certificate” advised Radoslav when explaining the AWS IoT Core service architecture.
Finally, he introduced the AWS IoT Greengrass, which extends AWS cloud to the edge, so that local compute resources can act autonomously and implement functionality as close to the IoT sources as possible. The service supports Lambda integration, local filtering, machine learning, security management and local messaging. Although the device management and the maintenance of all these features are empowered by the cloud, the Greengrass devices on the edge can operate offline. This results in significant cost saving combined with ease of operation and development.
At the meet-up, Radoslav Yovchev ran a short demo where he registered a new IoT device or “thing”. Then, he attached security certificates and configured the policies that allow the device to publish messages to the cloud. Once, the configuration was in place, he used his own laptop to publish MQTT messages via a python script to the MQTT message broker in AWS. Finally, he created an automation rule that took the messages and stored them automatically in a DynamoDB table.