Motion sensing detects that something is moving nearby, without identifying what or who. A PIR sensor, an ultrasonic pulse or simple camera-based motion detection is enough, and none of them recognise a person.
Two uses justify it on their own. Waking a screen when someone approaches and sleeping it when the space empties cuts energy use and panel wear substantially in places that are busy for only part of the day. And triggering an attract sequence at the moment someone arrives is far more effective than a loop that has already been running for six minutes.
The caveat is placement, and it is where most deployments go wrong. A sensor covering a corridor triggers on everyone walking past to somewhere else, so the screen is permanently awake and the saving never materialises. Aim for the zone where someone has actually stopped, not the one where they pass.
Worth noting for privacy reviews: motion sensing is not surveillance and should not be described as it. There is no image, no identity and nothing retained, which makes it a much easier deployment than anything camera-based.