In an IoT environment, data platforms need to be scalable.

As part of the IoT learning environment project, a general-purpose data platform is being developed for storing and analyzing environmental data collected by sensors. The aim is to use the same software architecture in all learning environments within the project, so particular attention must be paid to the adaptability and scalability of the platform.

An IoT system typically consists of numerous, even hundreds, of individual sensors and devices. These small devices have limited computing power, may come from multiple manufacturers, and may require manual configuration. Devices designed for different purposes may also use different formats or protocols for presenting and transmitting data. It is also important that new devices can be added to the system after its implementation if there is a need to expand the sensor network, for example, to monitor new measurable targets. Devices can be damaged in harsh conditions, so it should be possible to maintain and reconfigure them or replace them entirely with new ones. To avoid errors, the addition of new devices should be as automated as possible—manual editing of device identifiers and configuration data is not only laborious but also prone to human error.

In order for the collected measurement data to be of practical use, it must, of course, be possible to analyze and visualize it. Different views of the data must be created to meet the needs of different user groups, while at the same time ensuring secure access control to the data.

The use of cloud-based services speeds up the development of applications designed for data collection and utilization. Due to the heterogeneity of sensor network devices, services often need to be supplemented with various customizations, the implementation of which must pay close attention to scalability and future expansion needs. Accurate mapping of the data structure enables reliable data transfer and interaction between different services.