AeroDash: A Free and Open-Source Dashboard for City-Level Air Quality

The increasing volume and heterogeneity of data generated by IoT devices for environmental monitoring pose significant challenges for data integration, interpretation, and decision support.

Most existing platforms function primarily as data dashboards, focusing on raw data visualisation with limited analytical intelligence. To address these limitations, researchers at Newcastle University from across the NU-DICE lab and the Newcastle Urban Observatory developed AeroDash, an AI-driven visual analytics framework for urban AQI monitoring. AeroDash integrates real-time data fusion, hybrid ML-based forecasting, and interactive visualisation within a modular and scalable architecture that supports transparent/customisable dashboards. The framework incorporates predictive models, including linear regression, Holt–Winters, XGBoost, and long short-term memory (LSTM) networks, to capture linear-nonlinear relationships and temporal dynamics in pollutant behaviour.

The system is evaluated using live sensor data from a large-scale urban observatory and includes an automated self-validation mechanism that continuously monitors model performance. When prediction errors fall below 20%, models are automatically updated within the open-source AeroDash platform, enabling intelligent, adaptive, interpretable decision support for both expert and non-expert users.

This system will soon be integrated within the Urban Observatory, one of the world largest and leading urban observatory networks dedicated to monitoring, analysing, and providing open-access data on urban indicators for research and evidence-based decision-making.

The NU-DICE Lab team involved in the project included: Dr Sneha Verma, Balendra Vir Singh Chauhan, Xiang Xie and Prof. Mohammad Kassem, and the Newcastle Urban Observatory (Prof. Philip James).

AeroDash: Demonstration

Newcastle University

Last modified: Wed, 25 Feb 2026 13:06:02 GMT