Studentprosjektforslag - Vibration data acquisition and analysis for predictive maintenance

Sverres home


Introduksjon

Eksterne oppgaver:

Post-sorteringssystem

Vibration data acquisition and analysis for predictive maintenance

Ultra low power accelerometer

Secure Edge support for end-to-end encrypted sensor to cloud protocol

IoT sensor data authenticity and protection

ML-analysis on miniature sensors

Ultra low power Analog Sensor Interfaces

DNV Fuel Fighter - Datavisualisering av telemetridata

Ntnu Cyborg (flere oppgaver)

SW development projects:

Programmeringsprosjekt (Flere oppgaver)

'Big data' analysis of flash-card learning data

Procedural Generation: Game Worlds

Procedural Generation: Music

Programvare design av simulator

Dynamic deployment system for real-time tasks

Measurement-based real-time system

Bibliotek for meldingssending

Teoretical projects:

Schedulability proof for message passing systems

Dynamic deployment system for real-time tasks

Bruk av online eksekveringstidsestimater

Real-time systems not based on timing requirements

Morsomme sensorer og applikasjoner:

Døvehørsel

Blindesyn

Vibration data acquisition and analysis for predictive maintenance

New ultra low power accelerometers have become available in recent years. These devices can be used to measure vibration on various types of equipment, in order to predict mechanical breakdown at an earlier stage than what would be possible using other methods such as temperature sensing. Methods to capture and analyze vibration patterns with minimal energy consumption should be investigated. A demonstrator should be built based on an evaluation board of a suitable accelerometer.

The assignment will be to:

  • Assess methods to analyze vibration with respect to the tradeoff between energy consumption and accuracy.
  • Investigate the degree to which useful predictions can be reconstructed from partial data (such as an undersampled accelerometer signal). Address whether the required signal processing can be performed locally on the sensor's microprocessor, or whether work should be offloaded to the cloud.
  • Survey suitable accelerometer ICs.
  • Plan the architecture and design of a prototype device .

The assignment is well suited for continuation into a Master Project.

Feel free to make direct contact with the Disruptive Technologies external supervisor under it this sounds interesting.

Supervisor: Jonathan Reichelt Gjertsen, Disruptive Technologies jonathan.gjertsen@disruptive-technologies.com

Editor: Associate Professor Sverre Hendseth Contact Address: Sverre.Hendseth...ntnu.no Last Modified: 29/4-2020