People sharing one space have different needs

What it is

Ambient devices act on physical quantities shared by the space — a room has exactly one temperature, one brightness, one volume — while the people in it differ in preferences and activities. This is the structural fact of shared control: the divergence exists the moment more than one person is present, and no single setting satisfies everyone. It is not a symptom of poor user onboarding.

It is also not "several people sharing one device". Taking turns on the TV is a time-allocation problem; heating, lighting, and ambient sound act on everyone present simultaneously — there is no turn to take. The conflict is concurrent and continuous.

Why it happens

Three layers of divergence. Physiological: thermal comfort varies with metabolic rate, clothing, age, and body composition — one person's chilly is another's stuffy, and neither is being difficult. Activity-based: at the same moment one person reads (wants bright and quiet), another watches TV (wants dim and loud), a third sleeps (wants dark and silent) — the requirements clash directly. Temporal: household schedules stagger who is home when, so "the home's temperature" is really several populations' needs taking shifts in time.

One level deeper: there is no fair aggregation function. Averaging two target temperatures yields a temperature neither finds comfortable; majority rule makes minority members permanent losers; first-come-first-served delegates arbitration to the clock. Control over a shared quantity is an allocation problem, not an optimisation problem — which is why no amount of cleverness in a single-valued controller resolves it.

Studying it

• Domestic field studies: Davidoff et al.'s study of dual-income families (UbiComp 2006) found the home does not fit a single-controller model — multiple "home managers" with distinct routines and preferences; they argued control should shift among members and automation should keep manual override. Crabtree and Rodden's ethnography of domestic routines (CSCW 2004) likewise shows ambient device use embedded in multi-person coordination.
• Measuring co-presence disagreement: thermal comfort research records simultaneous comfort self-reports from co-located occupants (questionnaires or experience sampling), yielding a "disagreement rate" — how often two people in the same room at the same time report opposite comfort directions.
• Logs plus diaries: control logs (who adjusted what, when) combined with household diaries reconstruct how divergences arose and were settled.

One methodological caution: single-occupant lab comfort studies cannot produce this phenomenon — disagreement requires co-presence. Evaluating multi-user ambient control has to happen in genuinely shared living spaces.

Where it stops holding

• Applies only to systems acting on shared physical quantities (heating, lighting, ambient sound, ventilation). Personal-output devices (headphones, personal screens, wearables) are out of scope; physically zoned systems (grouped lighting, per-room climate) split the shared quantity back into local ones, defusing much of the structural conflict.
• Intensity rises with spatial density: shared dorm rooms and multi-occupant bedrooms are sharpest; in large homes each zone governs itself, and conflict concentrates in unowned common areas (living room, kitchen).
• "Whose preference should win" is a question of social order (seniority, ownership, who moved in first). A technical solution that ignores the household's existing settlement — defaulting, say, to the younger generation's energy-saving preference over an older member's warmth — is experienced as overreach, however "optimal" the numbers.

Applying it

• Treat "several people, one room" as the default operating condition, not an edge case: from the first version, the control surface should express whom the current setting serves, not assume a single user.
• Zone wherever zoning is possible: split one shared quantity into local ones (light groups, zoned climate, directional sound) — more effective than any arbitration algorithm.
• Where a single output is unavoidable, surface the divergence: show the gap between the current setting and each member's preset instead of hiding it inside a "smart" default.
• How to check: track per-member override rates — how often each person changes the setting. One member overriding persistently and frequently means their preference is losing structurally; that signal surfaces problems earlier than satisfaction questionnaires.

Related

• Same group: Z6.01.2 Last-write-wins overwrites turn into disputes · Z6.01.3 Conflicts need negotiation support, not silent overwrites
• Nearby: Z6.02 Permission layering · Z4.07 Multi-user and permission assignment
• Search terms: shared control · multi-user smart home · thermal comfort · preference conflict