Spatial legibility in hospital corridors: what robots announce and what they conceal.

A robot that occupies space communicates something about that space — whether its designers intended it to or not. Three observations from healthcare deployments where the signal was wrong.

First observation: a mid-sized autonomous delivery robot operating in a 1.2-meter corridor communicates, by its presence alone, that the corridor is now a shared lane. Staff interpreted this as the hospital's approval of a new routing norm. It was not. The robot's physical footprint had re-signaled the space without any organizational decision to do so. Spatial Legibility includes what the robot implies, not only what it displays.

Second observation: a robot with a flat top surface and no visual barrier was routinely used by staff as an impromptu shelf for lightweight items — cups, folders, gloves. The robot's form factor communicated "available surface" with more force than any procedural memo communicated "do not place items on the robot." Design encodes behavioral permission. If the shape invites the wrong behavior, the shape must change.

Third observation: a robot with a consistent patrol route began to function as a spatial anchor — staff timed handoffs and brief meetings by its passing. When the route was changed without notice, the disruption was reported as "the schedule is broken." The robot had become infrastructure. Infrastructure changes require communication. The team had treated it as a device update.

The implication across all three observations is the same: Spatial Legibility is not a feature of the robot in isolation. It is a property of the robot in its specific environment, read by the specific people who share that environment every day. Auditing Spatial Legibility means studying how people interpret the robot, not how the robot was designed to be interpreted.