FunSNM Work Package 1 Tutorial: Data Quality and Uncertainty in Sensor Networks
This tutorial presents a summary of Work Package 1 of the FunSNM project, focusing on data quality and uncertainty in sensor networks from a metrological perspective. It addresses key challenges in modern sensor networks, particularly those arising in IoT, large‑scale deployments, and resource‑constrained sensing environments.
The tutorial introduces core data quality concepts, including data quality characteristics, measures and metrics, and the role of metadata in improving trust, reliability, and reusability of sensor data. It discusses why traditional calibration methods are often impractical for sensor networks and outlines a range of alternative calibration and data‑driven approaches.
A central element is the identification and discussion of data quality characteristics most relevant to metrology, based on recent international standards. The tutorial explains how selected characteristics can be interpreted, assessed, and improved in sensor networks, even when reference sensors or full traceability chains are unavailable.
Finally, the tutorial brings these elements together in a structured, standards‑based data quality management process, illustrating how data quality can be systematically defined, monitored, assessed, and continuously improved throughout the sensor data lifecycle.
What you will learn
- Key data quality challenges in sensor networks and IoT systems
- How data quality characteristics and metrics support metrological assessment
- Which data quality aspects are essential from a metrological perspective
- How alternative calibration and data‑driven methods support uncertainty and traceability
- How to apply a structured process for continuous data quality improvement
Who is this tutorial for
This tutorial is relevant for researchers, engineers, and practitioners working with sensor networks, IoT systems, and sensor data, as well as metrologists, data quality specialists, and students seeking an introduction to metrology‑informed approaches to sensor data quality.
This tutorial presents a summary of Work Package 1 of the FunSNM project, focusing on data quality and uncertainty in sensor networks from a metrological perspective. It addresses key challenges faced in modern sensor networks and IoT systems, such as large data volumes, limited access to sensors, and the impracticality of traditional laboratory calibration.
The tutorial introduces essential data quality concepts, including data quality characteristics, measures and metrics, and the role of metadata in improving trust, reliability, and reuse of sensor data. It discusses alternative calibration and data‑driven approaches that can support uncertainty assessment and traceability when reference sensors or strict calibration procedures are unavailable.
Based on recent international standards, the tutorial identifies and explains the data quality characteristics most relevant to metrology. These concepts are brought together in a structured, standards‑based data quality management process, illustrating how data quality can be defined, monitored, assessed, and continuously improved throughout the sensor data lifecycle.
The tutorial is aimed at researchers, engineers, metrologists, and practitioners working with sensor networks, IoT systems, and sensor data analytics, as well as students seeking an introduction to metrology‑informed.
Watch the video here:
FunSNM Work Package 1 Tutorial - Data Quality and Uncertainty in Sensor Networks