Utility Perimeters
Use fixed views that include both the fence line and the ground below it.
A missing section, fallen panel, or open gap can change a perimeter’s condition. StreamVLM™ can route that visible condition to an operator for inspection.
The Visible Condition
The condition in scope is a missing, broken, displaced, or visibly open section of a fence or barrier. It is a visible scene detail, not a diagnosis, a conclusion about intent, or an automatic decision.
In StreamVLM™, an authorized operator writes a prompt such as “Alert when a perimeter fence has a visible gap.” The model evaluates selected live frames against that prompt and sends a match, camera, location, and time into the IREX event workflow for human review.
StreamVLM™ runs on a GPU node in the customer instance. Each prompt has its own confidence threshold, alert cooldown, event type, and camera selection.
Explore StreamVLM Prompt Detectors →Where It Helps
Use fixed views that include both the fence line and the ground below it.
Compare the alert with the live image and prior inspection records where available.
Create a maintenance or security task only after a person confirms the condition.
Illustrative Camera Frames
These illustrative frames show the kind of visible condition a prompt can surface. They are not real incidents, deployments, or proof of detection accuracy.
A missing panel creates a clear opening. Staff inspect the actual perimeter and check the repair schedule.
From Prompt to Action
Write a concise prompt for a missing, broken, displaced, or visibly open section of a fence or barrier, then select the cameras and times where the condition warrants attention.
Open the event with its camera, location, time, and available video context. Confirm what is visible before treating the match as actionable.
The authorized team applies the local procedure. StreamVLM supplies a review cue; it does not decide the response.
Limits to Test
No. The prompt concerns the visible state of the barrier. It does not determine cause, time of damage, or responsibility.
No. It reports a match to the visible condition in the prompt. An operator reviews the frame and local context before deciding what it means.
Start with representative cameras, define what makes a useful alert, then adjust the camera selection, confidence threshold, and alert cooldown during a camera-specific pilot.
IREX does not publish a universal accuracy figure for this prompt. Image quality, camera angle, lighting, occlusion, and the local definition of a useful alert all affect results.
Pilot on one fence run with a clear baseline view and a named owner for inspection and repair decisions.