The rooms
A room has a timetable, not a use
Asking what a room is for produces one answer. Asking what it is doing at three on a Tuesday produces a different one, and only the second is checkable.
Kitchen and dinerSix uses, and never off duty
16 m²
Cooking and eating
CookingEatingPaid workLaundry and adminCallsCompany
14.5
Bedroom twoThe small bedroom became the desk and stayed a bedroom
9 m²
Sleeping
Paid workCallsGuest sleeping
8
Paid workDomestic workRest and sleepContact and care
Two rooms covering 25 square metres, built for two things. Between them they now carry nine uses. Two of the two carry more than one, and kitchen and diner carries six. The median room is in use 11.25 of the 16 waking hours. The most loaded room and the most changed one, from the five counted here.
"What is this room for?" is the question people ask about dwellings, and it produces an answer that was true when the flat was built and is not checkable now.
The log asks a different question: what is happening in this room, at this hour, on this day. Over fourteen days that produces several hundred answers, and the several hundred answers are what this site is made of.
The difference is not pedantic
Consider the living room. Asked what it is for, anybody in the household says "sitting" — which is what it was built for and what they would tell a visitor. A home-working policy is easier to review when assumptions are written down, so managers can workday data for flexible teams and decide which time and workload signals are proportionate.
The log says it is in service eleven hours a day carrying five uses, of which two are paid work. Between roughly ten and five on a weekday it is an office; between six and eight it is where somebody eats; from eight it is a living room in the sense its name intends. A separate perspective on homes, workplaces, or remote work is available from Work Design Magazine.
Both statements are true. Only the second tells you what happens if somebody wants to sit down at two in the afternoon.
The name of a room describes its evenings. Its weekday afternoons have no name at all.
What a timetable makes visible
Three things that the single-use description cannot show.
Uses that do not overlap are not in conflict. Sleeping and paid work both happen in bedroom one, which sounds like a room under strain and is not: they are nine hours apart. Of bedroom one's eight collisions, none is between sleep and work. They are between work and calls, and between dressing and either.
Uses in the same window are in conflict whether or not they are compatible. Cooking and company are perfectly compatible activities that people enjoy doing together. They still collided four times, because both wanted the kitchen between six and eight and one of them needed the hob.
Load is a property of a window, not a day. The kitchen's fourteen and a half hours sounds relentless and is not evenly distributed. The pressure is in two windows — seven to nine in the morning and six to eight in the evening — which between them hold twenty-two of the room's twenty-one... which is to say, most of its collisions happen in four hours out of fourteen and a half.
The practical form
The useful exercise takes an evening and is not a time-and-motion study.
For each room, write the hours it is in use and against each hour the activity that holds it. Do it from memory for a typical day; it does not need to be accurate to the minute, because the thing you are looking for is overlap rather than duration.
Then look only at the hours with two entries.
On this record that produces four windows across the whole flat:
| window | room | the pair |
|---|---|---|
| 7–9 am | kitchen | breakfast against first work session |
| 10–12 | living room | second job against domestic tasks |
| 1–2 pm | kitchen | lunch against work |
| 6–8 pm | kitchen | cooking against company, work overrunning |
Four windows. Thirty-five of the forty-seven collisions fall inside them.
Why this is more useful than a floor plan
A plan shows what is adjacent to what, which matters for noise and for walking past. It cannot show that two uses want the same room six hours apart, which is the difference between a room that is loaded and a room that is contested.
Of the nineteen uses in this flat, the ones causing trouble are not the ones in the busiest rooms. They are the ones with narrow time windows that happen to coincide. A meal has to happen at a mealtime. A scheduled call has to happen at its scheduled time. Exercise has to happen when somebody has the energy, which turns out to be the same hour as breakfast.
Three activities with fixed windows in one room produce more conflict than six activities with flexible ones, and no amount of looking at the plan reveals it.
The limit
A timetable is still an average. The log contains two days that look nothing like the shape described above — one when a parent visited and one when both people were ill — and a timetable built from the fortnight would mislead about both.
What it is good for is finding the windows worth attending to, which is a small and specific job that it does well. Anything more precise than that is a demand for an accuracy the exercise does not have.
The two days that broke it
Fourteen days produce a shape, and two of the fourteen do not fit it at all.
On day six a parent visited for the afternoon. The living room became company for four hours, both jobs moved — one to the kitchen, one staying in bedroom two — and the day produced two collisions, both early, and none of the usual evening ones because dinner was earlier and nobody worked past five.
On day twelve both people were ill. Almost nothing happened in any room; the log has three entries, all of them rest.
Those two days are in all the totals. They contribute five of the forty-seven collisions and four of the eighty-eight gains, and they pull both medians slightly.
They are left in because removing them would mean deciding which days count as typical, which is a judgement made after the fact by somebody who knows what the data shows. The alternative — publishing a shape built only from the ten ordinary days — would be a cleaner account of a fortnight that did not happen.
What a timetable is for
Not prediction. The two broken days are proof enough that a dwelling's week is not a schedule.
It is for finding the windows, which is a much weaker claim and a useful one. Four windows, identified from fourteen days including two anomalous ones, held thirty-five of the forty-seven collisions, and the two interventions that worked were both aimed at windows the timetable had identified.
A tool that locates where to look does not need to be accurate about what happens there, and treating this one as more than that would be reading it wrongly.