Norwegian version of this page

Statistical and machine learning methods for Sensor Data

Sensor data are multidimensional streams of observations from various sensor systems. We work mainly on sensor systems in the maritime and industrial sector with a particular interest on fault/anomaly/change detection.

About the project

For maritime safety surveillance we develop new approaches based on the availability of large arrays of sensors, which monitor condition and performance of vessels, machinery, or power systems. Sensor data are becoming increasingly available on global ship fleets, with efficient broadband connectivity to shore. We suggest new generic approaches to condition and/or performance monitoring, which is the process of identifying changes in sensor data that are indicative of a developing anomaly or fault. In addition to using previous failure data and pattern recognition techniques to detect anomalies, we test model-based approaches that exploit knowledge on the sensors and the conditions they assess. We also rely on other data sources such as AIS data for the study of manoeuvres and collision avoidance.

Sub-projects

  • Sequential learning and decision based on sensor data from maritime (ship) data, PhD project for Fredrik Lundvall Wollbraaten
  • Data-driven state of health modelling for maritime battery systems, PhD project for Clara Bertinelli Salucci
  • Online contextual anomaly detection in multivariate data streams, PhD project for Per August Jarval Moen

  • Clustering and automatic labeling within time series of marine log messages, PhD project for Emanuele Gramuglia
  • Bearing fault monitoring, PhD project for Jaroslav Nowak
  • Scalable change and anomaly detection in cross-correlated data, PhD project for Martin Tveten (finished in 2021)
  • Data-driven methods for multiple sensor streams, with applications in the maritime industry. Phd project for Andreas Brandsæter (finished in 2020)
  • Realistic autonomous vessels test bed design using historical AIS traffic data and collision avoidance rules. Post doc project for Azzeddine Bakdi (finished in 2021)

Financing

Norwegian Research Council through the SFI BigInsight, dScience, ABB, DNV

Cooperation

Norwegian Computing Center, ABB, DNV

Published Sep. 29, 2022 8:55 AM - Last modified Sep. 29, 2022 2:05 PM

Participants

Detailed list of participants