Attendance may soon be measured, not taken

Under a pilot proposal circulating among the schools of one operator, attendance would no longer be something a teacher takes. It would be something the building measures. Corridor presence sensors and the wireless network would together produce a minute-by-minute record of where each pupil is, and a parent whose child is not where the model expects during the mid-morning break would receive a notification before the break ends.

From the staffroom this reads as an administrative improvement. From the third row it reads as the closing of a loophole that has been in continuous operation since the invention of the register. The old system asked a question once, at the start of the lesson, and then stopped asking. Everything a thirteen-year-old has ever arranged has been arranged in the part where it stopped asking.

The ten minutes matter more than the record does. There is a stairwell behind the gym where the signal drops, a bench past the bike sheds that no camera was ever pointed at, and a corridor by the old storeroom that is officially a fire route and practically another country. Under the pilot, the place where the signal drops would not be a gap in the data. It would be an event.

“We would not be tracking children,” a spokesperson for the operator would say. “We would be recording device associations, room occupancy and door events. Children happen to be standing next to all three. We consider that a meaningful distinction, and we accept that no thirteen-year-old has ever accepted it.”

The system would learn what a normal day looks like and flag the deviation, which is sensible design for a network and strange design for a person. A pupil who eats lunch alone in the library every day would be flagged on the day he sits down in the canteen. The alert would read unexpected location. It would not read finally.

Teachers consulted informally would raise something else. Calling the register is the least efficient thirty seconds of the school day and the only thirty seconds in which an adult looks at every child in the room in turn and hears each of them say one word out loud. The pilot would recover those thirty seconds. It does not say what would be done with them.

Nothing is installed anywhere. The document is a proposal with a costing annex and a section on parental consent that runs to a page and a half. The year group named in the consultation would already have worked out which corridors the sensors would go in, which lesson produces the most tolerated movement, and that a phone left in a bag in an empty classroom is, as far as the model is concerned, a child sitting quietly at a desk.