Identify areas for review
Where supported, automated tools can flag candidate anomalies for closer inspection. Agree which conditions the selected software is designed to identify before using its results.
AI-Assisted Inspection
AI-assisted inspection with human review.
Connect candidate findings with captured evidence and professional review. For MX-100 projects, confirm the available AI tools, supported defect classes, and validation for the selected software configuration.
Evidence and review
Make findings traceable from image to report.
Keep the supporting evidence
Retain the relevant image or model view alongside each finding. Record the location, reviewer notes, and measurements required by the inspection procedure.
Confirm findings with a reviewer
A qualified reviewer assesses the evidence, resolves ambiguous cases, and determines which findings belong in the final inspection record.
Match the software to the task
Confirm feature availability, licensing, defect categories, and reporting fields for the supplied configuration. A sample reconstruction alone does not establish detection performance.
How it works
From captured evidence to a reviewed record.
- 01
Capture
Collect interior imagery and scan data with the MX-100.
- 02
Reconstruct
Process the recording into the available spatial and visual outputs.
- 03
Review
Assess candidate findings against the images, geometry, and site context.
- 04
Document
Record confirmed observations and supporting evidence in the report.
This describes an inspection review workflow. Automated detection availability and scope depend on the selected software.
In practice
Review a continuous view of the manhole wall.
An unwrapped wall image helps a reviewer examine the interior surface and relate observations to the inspection record. The MX-100 page also shows imagery, mesh, point-cloud, and report examples.
Practical evaluation
Evaluate detection and documentation separately.
- Define the inspection scope. Agree the target conditions, evidence required, and reviewer responsibilities.
- Use representative recordings. Include the lighting, surface condition, obstructions, and manhole types expected on the project.
- Review both missed and incorrect flags. Compare the software’s findings with a reference review and document where human interpretation is needed.
- Check the delivered report. Confirm that observations, locations, images, and measurements transfer into the required template.
Questions about AI-assisted inspection
Does downloading the sample provide access to AI tools?
The sample is an MX-100 BAG source recording. Confirm processing software access, licensing, and any AI features with ModeXion. Review the MX-100 sample →
Which defect types are detected?
Detection categories and performance depend on the software configuration and validation. Ask for a demonstration using examples of the conditions your team needs to assess.
How does 3D data support inspection review?
Geometry and wall imagery provide complementary views for documenting observations and reviewing measurements. Understand point clouds and models →