Advancements in Fiber Optic Sensor Technology Pave the Way for Innovative Health Monitoring Systems

Strategic Insights -

A new report from a research team at the University of Aveiro in Portugal details the development of a novel dataset and approach to gait-based user identification using optical Fiber Bragg Grating (FBG) sensor-based accelerometers. The research aims to address the gap in previous works in this area, which were often limited by the lack of comprehensive datasets. The study found that by employing unsupervised machine learning models, such as K-means and DB-scan, it is possible to achieve high accuracy in distinguishing individuals.

Key Takeaways:

  • The research team developed a novel dataset and approach to gait-based user identification using a set of four optical FBG sensor-based accelerometers integrated into smart home environments.
  • The approach computes features such as entropy, mean, standard deviation, kurtosis, and skewness from raw signals, achieving high accuracy in distinguishing individuals.
  • The dataset results for each sensor were as follows: Sensor 1, K-means with a Silhouette Score of 0.3038 and Davies-Bouldin Index of 1.1238; Sensor 2, K-means with a Silhouette Score of 0.3543 and Davies-Bouldin Index of 0.9659; Sensor 3, DB-scan with a Silhouette Score of 0.2650 and Davies-Bouldin Index of 0.6855; and Sensor 4, K-means with a Silhouette Score of 0.3949 and Davies-Bouldin Index of 1.0438.
  • The approach not only enhances user identification but also facilitates personalized healthcare applications and unobtrusive monitoring.
  • The research has been peer-reviewed and published in the IEEE Transactions on Consumer Electronics journal.

Statistics:

  • The approach achieved high accuracy in distinguishing individuals using unsupervised machine learning models.
  • The dataset results for each sensor showed various Silhouette Scores ranging from 0.2650 to 0.3949 and Davies-Bouldin Index values ranging from 0.6855 to 1.0438.
  • The research found that employing unsupervised machine learning models, such as K-means and DB-scan, can lead to high accuracy in distinguishing individuals.

Sources:

  • Gait-based User Identification for Smart Home: When Machine Learning Meets Fbg Accelerometers, published in IEEE Transactions on Consumer Electronics (volume 71, issue 1, pp. 2332-2341, 2025).
  • Institute of Electrical and Electronics Engineers (IEEE) - www.ieee.org/ and IEEE Transactions on Consumer Electronics - ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=30.
  • University of Aveiro, Telecommunications Institute, P-3810193 Aveiro, Portugal.