Data Models standardization initiatives
- Onesait Platform
Introduction
One of the main benefits of data models is their reusability, not only in our own projects, but also with other clients and their systems.
Therefore, there are several initiatives that seek to generate data modeling standards in various fields, such as metadata management (ISO/IEC 11179), clinical data exchange (CDISC) or electronic health information (HL7), data and business process modeling (OMG) or the description of digital resources (Dublin Core).
In the world of the Internet of Things (IoT), standardization is essential to ensure the interoperability and efficiency of connected systems. Several initiatives have emerged to establish data models to ensure consistent and structured communication between IoT devices. However, none of them currently has sufficient market penetration to make a unequivocal choice.
Initiatives
Below is a table with relevant information on some of these initiatives:
Initiative | Description | Organization | Format | Version |
|---|---|---|---|---|
SensorML is a standard developed by the Open Geospatial Consortium (OGC) that provides an XML framework for describing sensors and their observations. This model allows the representation of sensor metadata, including their location, capabilities and calibrations. | OGC | XML+XSD | 2.1 (2020) | |
AMON is an ontology developed for the advanced management of metering data in smart grids. This initiative focuses on the representation of data from smart meters, facilitating their integration and analysis. | AMEE | JSON | 3.1 (2012) | |
The SensorThings API has been developed by the OGC, and is a standard for managing and exchanging sensor data on the web. This RESTful API enables real-time communication between IoT devices and applications, facilitating sensor data management. | OGC | JSON+ODATA | 1.1 (2020) | |
The GSMA IoT Big Data Harmonized Data Model is a GSMA initiative that provides a harmonized data model for the collection and analysis of large volumes of IoT data. This standardized model enables interoperability and efficient analysis of large-scale IoT data. | GSMA | JSON +Schema.org - Schema.org | 6.0. (2018) | |
https://www.fiware.org/data-models/ FIWARE is an open source platform that provides a set of standards and components for IoT application development. The FIWARE Data Model defines a set of standardized data models to represent information in various verticals, including smart cities, agriculture, energy and more. It adapts the GSMA Data Model by replacing Schema.org - Schema.org for Data Model definition with JSON-Schema to simplify its use. | FIWARE | FIWARE + JSON Schema | 1.0. 2017 | |
Smart Data Models is a joint initiative of FIWARE and other partners that seeks to develop a set of open and reusable data models for different domains. These data models are designed to be used in IoT and smart city applications, promoting interoperability and data reuse. | FIWARE Foundation + IUDX + TM Forum + OASC | JSON Schema + export in YAML, SQL, and soon in RDF | 1.0 (2020) |
What do these initiatives have in common?
Most of these initiatives (especially the most current ones) use JSON as the information exchange format. Regarding the JSON format, we can remark that:
It is a lightweight interchange format, which has now become the standard for data interchange, replacing XML format in most scenarios. An example of the difference between XML and JSON would be the following:
Source: WikipediaIt is the format used for everything by the big companies in the web world, such as Google, Amazon, etc.
It is the default format of the “API Economy” with which, for example, banks and organizations offer their data.
It is a format that is used in modern web and mobile applications, as it is a much lighter format than XML.
When storing and consuming data with a spatial component, the GeoJSON specification uses the JSON format.
It is a lightweight format, well suited for IoT devices.
To define the structure of a JSON file (for example, which attributes are mandatory or the data type of each property) there is JSON-Schema, the equivalent of XML-Schema, which is currently widely used, although it is neither mandatory nor standardized, as there are other initiatives.
About JSON and JSON-Schema and below standards such as those mentioned above, we find other formats and definitions such as OData or http://schema.org , but neither have a massive penetration for the moment.