In today’s rapidly changing world, the need for trustworthy measurements has never been greater. Sensor networks are everywhere, from monitoring city air quality and managing energy grids, to ensuring efficient manufacturing and smart building management. Yet, these networks face significant challenges: data quality is often unknown, uncertainties remain unaccounted for, and traceability to the International System of Units (SI) can be difficult to establish. As the digital transformation accelerates and the use of machine learning over the stream of data becomes routine, the reliability of measurement data becomes a central issue for industry, policymakers, and society.
Accurate measurement is the cornerstone of innovation and safety. As sensor networks expand across critical infrastructure and the environment, ensuring their outputs are reliable and traceable is vital. Without robust metrological assessment, we risk making decisions based on faulty or inconsistent data, which can impact everything from energy efficiency to public health. Recent European regulations also demand higher standards of data quality, traceability, and uncertainty management in sectors like air quality, industrial emissions, and energy efficiency.
The VTT coordinated project, FunSNM, is progressing well to address these urgent needs. Our goal is to establish practical, standardized methods for evaluating the quality and uncertainty of measurements from sensor networks, including distributed and dynamic systems. We’re developing new tools for uncertainty propagation, data quality metrics, and SI-traceability, and demonstrating them in real-world use cases across industry and public services.
But FunSNM goes further. We are applying machine learning to sensor networks, enabling smarter calibration, drift detection systems, and real-time decision support. By exploring the potential of these digital tools, we’re helping lay the groundwork for the future of metrology, where measurement science keeps pace with fast-evolving technologies.
The project has already achieved several milestones:
- New methods and guidance for uncertainty analysis and data quality in sensor networks, including large-scale, distributed, and time-varying systems.
- Demonstrations in real-world applications: We’re applying our methods to city-wide air quality monitoring, district heating networks, and smart buildings, showing concrete improvements in measurement reliability and efficiency.
- Practical tools for industry: Automated calibration, in-situ sensor drift detection, and uncertainty-aware sensor fusion, all designed with end-users in mind.
- AI and ML integration: We are pioneering the use of machine learning for automated calibration and data fusion in a metrologically sound manner, enhancing both the performance and trustworthiness of sensor networks.
- Stakeholder engagement and impact: Our results are already informing European and international standards, and our open-access publications and software ensure broad benefits for industry, researchers, and the public.
Sensor Network Metrology is a practical example of how systems metrology is evolving. By combining rigorous science with advanced digital tools, we’re preparing measurement science for tomorrow’s challenges. As a result of our new developments, VTT MIKES empowers customers and stakeholders to make better decisions, improve efficiency, and build sustainable solutions for the future. Above all, the main goal of metrology remains clear: establishing trustworthy measurements that support innovation, safety, and quality of life for everyone.
For more information, publications, and updates, visit: www.funsnm.eu
