Magnetic fields for fault detection
We have developed magneto-resistive sensor testing and data analytics to monitor and detect faults in powerlines and inaccessible steel infrastructure.
The science
Induced currents in encased steel and operational currents in powerlines both produce magnetic fields which can be interrogated to detect defects and faults.
We develop and use magneto-resistive sensors to measure the magnetic field; either those induced by an eddy current, or those occurring around electrical conductors (in the case of powerlines). The magnetic signatures are then interpreted via a set of data analytics algorithms designed to identify the data patterns.
In the case of powerlines, the pattern recognition system has been built using machine learning to look at fluctuating magnetic fields due to changes in currents, and recognise anomalies.
Our current work also extends to using sensors and artificial intelligence-based control to mitigate potential hazards and ensure the safety of autonomous robots in rugged outdoor environments.
Impact and potential
Whether it is used in pipes or to reinforce concrete columns—such as in buildings, bridges, and power poles—steel eventually corrodes. Where that steel is underneath other material, whether underground or encased in concrete, it cannot be inspected by eye.
A non-destructive testing tool has been developed to identify changes in steel thickness and allow the asset owners to target maintenance effectively. This saves maintenance costs and prolongs the life of assets. Accurate assessment and timely maintenance can also prevent catastrophic failure occurring.
In the case of electricity distribution, we have developed a low-cost, widely distributed monitoring solution for current levels in powerlines. It allows asset owners to identify when and where faults are occurring. They are then able to avoid lengthy outages and unnecessary repairs, and target only those pieces of equipment that are showing signs of a fault.
Capabilities
- Expertise in developing sensing hardware
- Creation of sensor data interpretation algorithms
- Using artificial intelligence and machine learning in sensing and control systems
- Experience with advanced process control
- Knowledge and expertise in magnetic fields and their interpretation
The people
Principal engineer Dr Fiona Stevens McFadden likes to work in applied research and find solutions to real-world problems.
Principal Engineer
Robinson Research Institute