A cross-domain workbench for building, explaining and — critically — validating AI models before anyone trusts them. It couples model development with explainability outputs and ...

An end-to-end environment for developing and interrogating deep learning models on nuclear medicine and radiological images. It handles study import, model training and inference, and ...

A screening-oriented version of the suite, adapted to ophthalmic imaging. It supports automated detection of ocular disease from retinal images, presents case-level explanation maps alongside ...

A dedicated tool for turning raw low-cost sensor output into reference-comparable data. It ingests sensor and co-located reference measurements, applies machine-learning correction models on top ...

The current generation of the Smart Air Quality platform, extending measurement into forecasting. Alongside real-time monitoring, it produces short-term pollutant predictions from the network’s own ...

The intermediate generation of the platform: a consolidated monitoring system combining multi-pollutant low-cost sensing, automatic correction of the raw signals, and continuous delivery of quality-checked ...

The compact member of the ENSENSIA family — a smaller, lower-cost and easier-to-deploy unit built for dense placement and indoor or personal-scale use, while remaining ...

The flagship air quality monitoring node and the commercial face of the work. A self-contained multi-pollutant station that measures, corrects and transmits continuously, integrating sensing ...

The overall architecture that ties the individual pieces together: sensing nodes in the field, ingestion and storage, correction and modelling services, and the user-facing dashboards ...

The administrative backbone of the network. A single interface for registering devices, tracking their configuration, location and health, and spotting units that have gone silent ...

The main monitoring interface: live and historical pollutant readings, per-device time series, comparison between units, and export of the underlying data. Built so that both ...

An analysis environment sitting on top of the stored data. It lets the user define experiments over selected devices and time windows, apply and compare ...

A geographic view of the monitoring network. Every station appears at its actual location with its latest status and readings, and can be opened directly ...

Made in the framework of the project PANSEN. A spatially continuous view of air pollution, produced by interpolating the network’s measurements into a heatmap over ...