Users and bystanders require separate consideration

What it is

The bystander–user distinction separates people who operate or receive a service from those affected merely because they share space with a robot. Bystanders may be recorded, displaced, exposed to sound, or confused by motion without training, terms of use, or access to controls. Treating everyone as a user hides this asymmetry in information, choice, and risk.

Why it happens

Users usually know the task goal and can interpret an anomaly as system state; bystanders infer intent from visible motion and public conventions alone. A shortest path, following distance, or spoken prompt optimised for a primary user can transfer cost to others. Each encounter may be brief, but changing crowds create a cumulative impact larger than that experienced by one operator. Direct-use performance is therefore an incomplete measure of public acceptability.

Studying it

Field work should distinguish operators, service recipients, and bystanders while measuring avoidance, pausing, detours, gaze, complaints, privacy perception, and comprehension of yielding signals. Video coding needs to separate intentional engagement from incidental interaction and avoid sampling only people willing to approach. Research recording in public also requires data minimisation for people who did not consent to participate.

Where it stops holding

Roles can change during an encounter: a bystander may be recruited to help, and a user may become passively exposed after the task. A low-traffic closed pilot does not represent a dense public space, while visible curiosity does not waive rights to passage, quiet, or non-collection. Bystanders need not receive a full manual, but they do need to judge approach and avoidance.

Applying it

• For each role, list benefit, burden, available information, refusal path, and emergency control.
• Put motion intent, stopped state, and safe passing space on the robot or in the environment, not only on an operator display.
• Under realistic traffic, count detours, emergency stops, blockages, and complaints per encounter and interview people who did not engage. If user efficiency increases by shifting burden to bystanders, revise routing and priority.

Related

• Same group: X1.03.1 Design begins by defining the task and the people who share it · X1.03.2 Structured and open environments impose different requirements
• Adjacent: X5.02 Movement in shared space · X6.06 Robot etiquette in public space
• Search terms: robot bystanders · incidental interaction · public-space HRI