AI-based near-miss monitoring uses cameras, and sometimes tags on trucks and people, to record close calls between forklifts, pedestrians and objects automatically instead of relying on workers to report them. Each event is logged with a time, a location and often a short video clip, and can be graded by how close and how fast the interaction was. Because many close calls are never reported, automatic capture shows where and when risk concentrates, such as one intersection on the night shift, so the layout, traffic rules or coaching can be changed before an injury occurs.
The full guide is Forklift near miss reporting, which covers the reporting program that automatic capture feeds. Systems differ in where they look. Analytics running on existing ceiling cameras see the whole floor but usually report after the event; cameras or tags on the truck see close range and can alert the driver at the moment. Before relying on the counts, check how the system defines a near miss (distance, speed or time to contact), how many flagged events a reviewer confirms, and who can see the clips, which the privacy guide covers. Track trends per operating hour rather than as raw counts.