Work at height on a lift carries risks that are easy to miss from the ground.
By the Sentrick Industrial team · Published October 25, 2026
Scissor lifts and boom lifts, often grouped as mobile elevated work platforms or MEWPs, make work at height faster and safer than ladders in many situations. They also bring their own hazards: tip-overs, collisions, falls from the platform, people trapped against structures and contact with overhead obstacles or power lines. Because these risks develop in seconds and are easy to miss from the ground, many safety teams are asking whether vision AI can help. The honest answer is yes in some areas, and not at all in others.
Cameras with computer vision can look at the things that are visible and repeatable. Examples include whether people are standing inside a marked exclusion zone around the base of a lift, whether the platform is being driven while elevated in an area where that is not allowed, whether a worker on the platform is wearing a harness and hard hat, and whether another vehicle or person is moving too close. In a well-configured setup these can become alerts to a supervisor or to the operator's ground spotter.
Vision AI can also support the work around the lift. It can flag blocked travel paths, obstructed emergency exits near the work area or equipment left in a walkway, which are the kinds of conditions that make a lift movement more hazardous.
A camera cannot measure ground bearing capacity, wind speed at the platform or the internal condition of the machine. It cannot confirm that the pre-use inspection was done properly or that the operator has been trained and authorized. It may also struggle with overhead power lines, which are thin, hard to see and require a proper site survey, not a camera. These limits matter: a system that appears to cover MEWP safety can create false confidence if teams forget what it cannot see.
Lift locations change as work moves, so fixed cameras alone may not cover every task. Options include fixed cameras on key work areas and corridors, mobile or relocatable cameras on poles or tripods for temporary tasks, and clear rules about when an area needs camera coverage and when it does not. Consider the angle: an elevated platform viewed from far below can be difficult to interpret, so aim for views that show both the base and the surrounding zone.
Keep alerts specific and actionable. An alert such as "person inside lift exclusion zone" is clearer than a general warning. Route it to the person who can act, such as the spotter or area supervisor, and pair it with a defined response: stop movement, clear the zone, confirm safe conditions. Tune the system with real footage from your site so lighting, shadows and reflective equipment do not create constant false alarms.
Vision AI should sit alongside, not replace, the controls that matter most. These include operator training and authorization, daily inspections, site-specific rescue plans, fall protection rules, ground assessment before setup and clear communication between the platform and the ground. Review alert data in toolbox talks to see where lift work creates the most near-misses, then adjust layout, procedures or supervision accordingly.
Begin with one or two lift tasks where exclusion zones and harness use are clear and measurable. Run the system in review mode for a few weeks, check the alerts against what supervisors saw and expand only when the results are trustworthy. Used this way, vision AI becomes another set of eyes for the people on the ground, helping teams notice problems earlier without pretending to replace judgment, training or inspection.