Construction site safety score: grade your site from a photo
A construction site safety score is a 0-100 grade of the visible conditions in a job-site photo. Upload a picture of any work area and the AI identifies each visible hazard — open edges, ladder setup, housekeeping, PPE, electrical — weighs them by severity, and returns the score with a finding-by-finding breakdown referencing the OSHA standard each hazard may relate to. Scanning the same areas week over week turns the score into a trend you can put in front of a GC, an owner, or an insurer. It is free to try with no account, and it is an AI-assisted snapshot of what the camera can see — not a formal compliance rating.
- Close every open edge
- Fix ladder setup everywhere
- Clear the walkways daily
- Cap or guard protruding rebar
How is the site safety score calculated?
The AI identifies each visible hazard in the photo, weighs it by severity — imminent-danger findings like open edges pull the score down far more than low-risk housekeeping items — and grades the frame 0-100. Every score comes with the finding list behind it, each referenced to the OSHA standard it may relate to, so the number is always traceable to specifics.
Is the score an official OSHA rating?
No. OSHA does not issue site scores, and no photo tool can certify compliance. The score is an AI-assisted snapshot of visible conditions, useful for tracking, prioritizing, and reporting — a formal compliance determination still requires a qualified professional reviewing the site itself.
What is a good score?
Treat the trend as the target rather than a magic number. Scores in the 90s mean the camera found little to flag; scores in the 40s-50s usually mean at least one serious exposure is visible. The most useful pattern is the same areas scanned weekly, moving up — and staying up.
Can I score subcontractor work areas?
Yes — any area you can photograph, you can score, and every area is graded on the same standard. Many supervisors scan each sub's zone during the morning walk so the weekly safety conversation starts from the same evidence instead of competing impressions.