Radar Tyre Monitor System

This project focussed on trialling a radar sensing technology designed for in-situ monitoring of signs of damage in haul truck tyres. Although a stand-alone project in terms of its focus on a radar-based sensor technology, it built upon insights developed in projects C25034 and C27074, which examined the use of an x-ray sensor for offline inspection of tyres.

The primary goal of this project was to establish a proof-of-concept prototype system that could demonstrate the ability of a radar sensor to detect damage (both internal and external) in tyres fitted onto large mine vehicles such as haul trucks.

The project successfully achieved its primary goal, delivering reliable and repeatable detection results that identified damage in tyres under lab-controlled circumstances. The secondary goal of detecting similar radar-trace patterns in field-based tyre measurements was also achieved.

The prototype sensor is designed to mount on the body of a mine vehicle (haul truck or similar) to provide a continuous monitoring system. Unlike the x-ray sensor, this proposed radar system would allow for in-situ monitoring without the need for costly, time-consuming and potentially hazardous tyre removal for inspection. Instead, the system would continuously monitor the state of the tyres while they are in service. For the purposes of this project, a single sensor mounting arrangement was trialled, however a future, practical detection system would feature separate sensors for each tyre on the vehicle.

The project work was scoped in three distinct phases:

1. Radar sensor design and development using simulated tyre segments consisting of interleaved layers of rubber and other material to establish basic performance metrics.
2. Radar sensor and machine-learning algorithm testing under lab conditions, using large passenger vehicle tyres to build a comprehensive dataset of scans from (1) new tyres, (2) tyres with introduced surface damage, and (3) tyres with introduced sub-surface damage.
3. Field trials deploying the radar sensor on (1) large agricultural tyres featuring various forms of real-world damage in known distributions, and (2) a mine haul truck with tyres displaying significant wear, but unspecified internal structural integrity.

The following sensor trial outcomes were successfully achieved:

• Discrimination between simulated tyre segment structures consisting of multiple layers of dense synthetic rubber, and segments consisting of the synthetic rubber with a metallic strip between layers, simulating the presence of the steel radials found within modern industrial tyres.
• Discrimination, by means of a machine learning (ML) algorithm, between three classes of passenger tyre under controlled lab conditions: undamaged, externally damaged and internally damaged tyres.
• Extensive scanning, processing and qualitative results assessment of large agricultural tyres with varying types and degrees of real-world damage.
• Acquisition, processing and visualisation of scans of tyres on an operational Caterpillar 777 haul truck, together with qualitative assessment of the data to gauge its suitability for interrogation by the ML algorithms developed for the laboratory-controlled testing.

The outcomes of this project demonstrate that the radar sensor developed is a viable technology for the monitoring of mine vehicle tyres for signs of damage, both external and internal.