Skip to content
This is NOT an official site of the Government of Canada. Click here for the official AI registry.

Helpful Places

Government of Canada – DTPR AI Register

AI-Assisted Drone Visual Inspection for Bridges

Safety & Security · Research & Development

What it collects

About a place
Anonymized data
  • Live drone-captured video and imagery of bridge surfaces and structures, representing the physical environment of the inspection site. No personal information is collected.
About a measurement
Anonymized data
  • Pixel-level visual measurements from drone cameras used as input to defect detection models, sourced from live drone feeds and benchmarked against the CODEBRIM dataset.
Run by
National Research Council Canada (NRC)
Where
No fixed location
Kept
Not stated by the Helpful Places.
Shared with
Accountable organization
Your copy
You cannot see the data it holds about you. What you can do

What it is for

Safety & SecurityResearch & Development

This system uses artificial intelligence to automatically detect and highlight surface defects on bridges in real time, using footage captured by drones. It is designed to help government inspectors do their work more safely, quickly, and cost-effectively. The system is currently a proof of concept used by National Research Council Canada employees only, and no personal information is collected or processed.

What it collects and what happens to it

Data taken in

About a place
Anonymized data
  • Live drone-captured video and imagery of bridge surfaces and structures, representing the physical environment of the inspection site. No personal information is collected.
About a measurement
Anonymized data
  • Pixel-level visual measurements from drone cameras used as input to defect detection models, sourced from live drone feeds and benchmarked against the CODEBRIM dataset.

Processing

Computer Vision
  • Computer vision algorithms analyze drone video and imagery to detect, identify, and localize surface defects on bridge structures, feeding results to an augmented reality display.
Classification & Prediction
  • Machine learning classification models trained on the CODEBRIM dataset assign defect category labels (e.g., crack, corrosion, spalling) to detected regions in bridge imagery.

What it does

Sensing (Perceptive AI)
Human decides
  • The system senses raw visual imagery captured by drones and processes it to detect and identify surface defects, converting pixel data into structured detections that are highlighted in augmented reality for human inspectors.
Deciding (Analytical AI)
Human decides
  • The AI classifies and labels detected surface defects (e.g., cracks, spalling) from drone imagery, producing structured identifications that are presented to inspectors who make final assessment decisions.

Outputs

About a place
Anonymized data
  • Augmented reality overlays highlighting detected and identified surface defects on bridge structures, displayed to inspectors in real time. Outputs are location-annotated defect markers on the bridge imagery.
A recommendation or prediction
Anonymized data
  • The AI produces defect identification and classification results presented as advisory highlights to inspectors, who retain final judgment on condition assessment and any remediation decisions.

Run by

National Research Council Canada (NRC)
  • National Research Council Canada (NRC) is the federal department deploying and operating this AI-assisted bridge inspection proof-of-concept system. Primary users are GC employees within NRC.

Government of Canada AI Register — 2526-NRC-CNRC-011

Built by

Not stated by the Helpful Places.

Kept for

Not stated by the Helpful Places.

Shared with

Not available to me
  • The system processes structural imagery of bridges only, with no personal information collected. Members of the public have no data access mechanism, as this is an internal government inspection tool.
Available to the accountable organization
  • Inspection outputs and defect detection results are available to GC employees at the National Research Council Canada who operate the system.

Stored

Not stated by the Helpful Places.

How to read the colours

Can it identify you?

Anonymized data
Data about people with the link to who is broken. Stripped of identifiers, blurred, aggregated, or noised so this system can’t reasonably tie a record back to an individual.
Pseudonymous data
Each person’s data is tied to a token (hash, ID, template) that lets this system recognise the same person across events, but the token itself doesn’t reveal a name. Reidentification is possible with extra information.
Identifiable data
The data either contains a direct identifier (name, address, account name, recognisable face or voice, plate number) or carries a token this system uses to look up legal identity during processing.

Who completes the loop?

Human decides
This mode suggests; a person decides what to do next. The AI is always advisory — a human is in the loop on every decision. Example: a triage tool ranks cases for a clinician who chooses which to see first.
Human executes
This mode decides; a person carries out the result. Example: an optimizer plans the day’s trash-collection routes, and drivers run them.
Autonomous
This mode decides and acts on its own. No person reviews each decision or carries out the resulting action.

Definitions from the DTPR standard. Amber is about your data, violet about who decides. The fuller the shape and the deeper the colour, the more identifying the data or the less a person is involved.

What you can do

Ask about this system

Questions go to the Helpful Places, not the vendor.

Your rights

  • Right to Algorithmic TransparencyThis system is listed on the Government of Canada's public AI register. As no personal information is collected or processed, individual data rights do not apply to members of the public. GC employees using the system are informed it is an AI-assisted tool at proof-of-concept stage.

Risks and safeguards

  • Physical harmMissed or misclassified defects could lead to underestimation of structural risk, potentially contributing to delayed maintenance and unsafe bridge conditions.Safeguard: The system is advisory only — human inspectors review all AI outputs and make final determinations. The system is currently at proof-of-concept stage, limiting deployment scope. Training on the CODEBRIM dataset provides a validated defect taxonomy baseline.
  • Reputational harmFalse positive defect detections could trigger unnecessary and costly remediation, or misidentification of bridge conditions could affect public confidence in infrastructure safety programs.Safeguard: Human inspectors validate all AI outputs before any action is taken; the proof-of-concept status means no production deployment decisions are based solely on system outputs.