Change Detection and Convergence Monitoring FAQ

Yes, technically all Hovermap models are supported for this workflow. However, the higher-accuracy Hovermap ST-X improves the accuracy of the convergence monitoring result and is therefore the recommended product for this use case.

No. Only Hovermap scans are supported.

No. Convergence monitoring currently only supports comparing two Hovermap scans.

The bag files are required for both the reference scan and second scan because the SLAM algorithm is run on the raw data as part of the alignment process.

The Aura alignment algorithm searches for matching features between scans within the Max Distance parameter, which defaults to 0.5 m. The closer the rough alignment of scans, the higher the likelihood of a good final alignment.

The Aura alignment algorithm fixes the reference scan in place and aligns the second scan to it. This has the benefit of reducing drift in scans aligned for Convergence Monitoring.

Measurements in a small controlled environment showed an average error of approximately 12 mm. Accuracy varies depending on factors such as the size of the scanned area, features in the scan, speed of the scan, and density of scanned areas.

False changes will be detected if objects or people are in one scan and not in the other.

Examples of unrelated changes include the vent bag being inflated in one scan and not the other, changes on the drive floor, and changes created by vehicles knocking out chunks of the wall.

The alignment of the scans and the resulting distance values may be affected in areas with limited matching features. Small changes such as missing chunks in the wall typically do not significantly impact the alignment of the scans.

Georeferencing is important because it helps with pattern recognition where some structure is moving in relation to another, providing context to the stoping front and supporting the integration of lithology data. Aura does not yet support using GCPs as part of the convergence monitoring workflow. Support is planned for a future release. In the meantime, some clients using this solution bring the Aura output into Cloud Compare and use a 4-point method with spray-painted marks to georeference the data.

Viewing data outside of the drive can be easier for analysis in some cases. Some clients are familiar with Cloud Compare's MC32 workflow, which creates a single skin by taking an average of all the points and creating a plane. By contrast, Aura takes an averaged mesh to a scan with a thickness, which means Convergence is viewed from inside the tunnel. If outside-the-tunnel visualisation is required, bring the two aligned Aura Convergence Monitoring outputs into Cloud Compare and run MC32 there.