Decentralised and federated systems
Federated data architectures unite semi-autonomous, decentrally organised database systems, to make interoperability and information sharing possible. Completely decentralised architectures store data in personal data vaults. This can be particularly useful in health-related projects, for instance.
One example is BIBOPP, a science-based tool to improve your lifestyle and reduce your risk of disease. BIBOPP uses Solid, a specification that lets people store their data securely in decentralised data stores called Pods. Another example is the We Are ecosystem, that uses personal health data for innovation.
Graph databases
Graph databases visualise problems and relations between objects. Rather than using tables the way a relational database does, graph databases use nodes, properties, and edges to define and store data. This approach makes them very popular in fraud detection, social networks, and other applications.
For instance, graph databases were paramount in solving the Panama Papers. At VITO, we take graph databases to the next level, by also focusing on properties that change over time, such as spatial and temporal data. This helps us make more accurate predictions about for instance water in Flanders or crowd management at festivals, ski areas etc.
IoT platforms
The Internet of Things (IoT) consists of devices and systems that exchange data, allowing us to draw substantiated conclusions, based in reality. VITO develops and uses several platforms that gather, combine, and make use of this data to make real-life predictions and act upon them. The permanent data streams from the IoT open the way for smart applications. For example: Internet of Water Flanders.
A fine-grained measurement and high-frequency sensor network is deployed at selected locations across Flanders and will continuously measure indicators of water quality from 2023 onwards. This will facilitate a more dynamic and efficient water management and pave the way to a more robust water system.
Fair data
FAIR stands for Findable, Accessible, Interoperable, Reusable. It specifies a list of 15 principles that open research data should ideally comply with. The end goal is to make sure that data will stay online and available for as long as can be guaranteed. Our data experts set up a platform where our different research units can handle their data in a way that fits their needs, and easily share that data-to-data catalogues and other repositories.
This way of working is aligned as much as possible with initiatives elsewhere. One example is Terrascope, an online platform to make open-source satellite images easily accessible to all users, free of charge.
More on data science?
Contact Bart Buelens and see how data science can help your organisation move forward.
