Application of Time Series Database for IoT Smart City Platform

IoT devices are becoming more prevalent yearly due to their relatively low cost and high maintainability. Their use cases are often varied-from industry (e.g., devices intended for quality control) to their application in smart cities, where they can help supervise day-to-day operations. With the increase in popularity comes the need to effectively store, query and process the data produced by such devices. This problem can seem relatively easy at first glance, but often, the solutions are not as straightforward. In real-world applications, multiple device types are needed to accomplish each task. This typically results in large streams of incoming data, each with its own structure and data representation. Due to the large amounts of devices, it may be necessary to distribute the workload or use serverless computing to tackle fluctuations in incoming requests. This paper provides a study of some options that can be used to achieve this goal, focusing on the efficiency and ease of use of both SQL and NoSQL databases currently available regarding deployment on both local devices and serverless functions like AWS Lambda. Document or relational databases are often used to store data from such devices. However, features of Time Series databases could be used to improve the entire system's performance and reduce storage requirements. One of the systems that could benefit from this improvement is a Smart City system developed by Logimic. The cost of the proposed smart city platform is essential because a high running cost can result in the system being more expensive than the cost reduction from using efficient IoT devices.