Loupe360 logoLoupe360 logo

Data processing

Loupe360 ingests data from surveys, historic reports, and other sources to extract insights and standardise their format.

We harness machine learning to automate defect detection and classification, improving survey efficiency. Our automated data pipelines centralize your datasets to aid analysis and accessibility.

Defect bounding boxes visualised in 360 degrees

Our other services

Use any combination of our services to best suit your needs. Loupe360 was built to support data from a variety of sources to allow you to capture your own data. Please Enquire if you have any questions.

Why process your data with Loupe360?

Maximize the potential of your data through advanced processing

Reduced inspection times

Automated defect detection performs a surveyor's task of identifying and classifying defects in a fraction of the time.

Enables you to quickly receive "at-a-glance" metrics in our Dashboard following a survey. Use these to identify areas that require further investigation without any manual or repetitive processes.

Our work with NRHS suggests that Loupe360 will reduce their end-to-end tunnel inspection times from 60 to 6 days over a 40km section of tunnel.

Improved data management

Processing data into a standardised format following each survey and storing it in a central location enhances data management.

Organised information is easier to access, analyse, and interpret. Consistency is particularly valuable when comparing successive surveys to allow reliable comparisons between datasets.

Combined with the ability to upload historic datasets in the same format, then manage everything via our Dashboard, you will make the most out of your data.

What we process

We automatically detect defects from 360 imagery and ingest your historical reports into Loupe360

360 imagery

Our machine learning model is trained to identify defects in tunnel linings from 360 imagery. Imagery can be captured yourself or with our Data capture service:

  • Successfully identifies and classifies cracks, water ingress, deposits, and corrosion.
  • Can be trained to learn new defects.
  • Compatible with most tunnel linings.
  • Provides a baseline tunnel condition to highlight areas of interest.

We are continuously training and making improvements to the model as we receive more data and feedback. Our vision is to make its predictions more reliable than a human.

Defects automatically detected using machine learning

Historic datasets

Loupe360 can ingest your historical datasets, usually previous inspection reports, into a standardised format for use by the Loupe360 platform:

  • Flexible to suit your unique .xlsx and .csv structure.
  • Upload point-and-shoot images of defects.

Our work with Metrolinx relies on this functionality to monitor the condition of their tunnel during construction. It enables their site logs to be regularly uploaded to Loupe360 so they can be referred to once the tunnel is operational.

Interface to upload historic datasets to Loupe360Interface to upload historic datasets to Loupe360