For more than 25 years, NILU has developed software and sensors for environmental monitoring — from the first national air quality systems to the data centres that today hold atmospheric observations on behalf of the United Nations, the World Meteorological Organization and the European Space Agency.
Our expertise covers the entire chain: the sensors that collect measurements, the databases that store the data, the AI models that interpret and predict, and the services that bring results to authorities, researchers and the public. Few institutes cover this whole chain under one roof. It allows us to turn research into solutions our partners can put into everyday use.
Our strategic roadmap for digitalisation and artificial intelligence supports the green and digital transition: automating large data streams, improving forecasts, and helping society move from reacting to environmental problems to preventing them.
- We develop and operate digital systems for environmental data.
- We publish environmental data openly and according to recognised standards.
- We build capacity, so that partners can run the solutions we deliver themselves
Competence areas:
Artificial intelligence and machine learning
Machine learning used on its own or combined with physics-based models, with calibration, correction and data fusion as core parts of the work. Models run in operational systems, trained in NILU’s own AI-lab, and assessed for trustworthiness through the Z-Inspection® process.
Data management and open data
NILU is among the world’s leading institutions for storing atmospheric composition data, and leads the ACTRIS data centre for ACTRIS ERIC. EBAS, EVDC and ACTRIS are organised under NILU Centre for Atmospheric Data, open and free of charge.
Digital transformation and capacity building
From separate systems and manual routines to integrated digital services — a transition NILU has been through itself. RAVEN, our open-source reporting system, is in operational use in the Western Balkans, Malta, Luxembourg and Romania.
Digital services, visualisation and AI-supported decision-making
Maps, portals, dashboards, decision support and mobile apps, with open APIs so others can build on the data. Users are involved from the start, and large language models are being taken into use for decision support on NILU’s own infrastructure.
Smart and intelligent sensing systems
Measurement technology combined with on-board processing and AI, so sensors can perceive, analyse and react. Low-cost platforms, data loggers and units on vehicles and drones, all validated against reference instruments.
Infrastructure for data services
Systems built to outlast the projects that fund them — several have run for more than 20 years. Relational and NoSQL databases, cloud-native container platforms, and semantic models for data that crosses disciplines.