Every option Archi4D generates arrives with two things beside the drawing: an area schedule, and four code checks. Most people look at the drawing and skip both, which is a shame, because the numbers answer a question the drawing can't — is this plan actually any good?
Here is how to read them in about thirty seconds.
The area schedule
A room-by-room table of floor areas, with the total at the bottom and the circulation share called out.
Total floor area
Compare it to the target you set, not to what feels right. If the plan came back 15% over, that is 15% more building to pay for — and it usually shows up in one or two rooms that are more generous than you asked for, not spread evenly.
Circulation
The share of the floor area doing nothing but connecting rooms: entries, corridors, landings.
| Circulation | What it means |
|---|---|
| Under 12% | Efficient. The plan is mostly rooms. |
| 12–15% | Normal for a house with a hall. |
| Over 15% | The plan is spending floor area on corridor. |
On a 100 m² house, each point is a square metre you are heating, cladding, and walking through. Over 15% is not a failure, but it should be a decision — a long gallery can be worth it — rather than an accident.
Room-by-room
Scan for two things: a room that is much bigger than its neighbours without a reason, and a room that has quietly fallen below what it needs to work. A 2.4 m² bathroom is technically a bathroom and practically an argument.
The four code checks
The checks are computed from the geometry, in the jurisdiction set in your preferences.
Egress path length
Measured room-centre to room-centre along the door graph, from the deepest room to the entry, and flagged against a 20 m limit. A fail here is nearly always a plan that has grown a tail — a bedroom wing reached through the whole house.
Fix it by moving the entry, or giving the far end a second door.
Ceiling height
Reported at the 2 700 mm the generator assumes throughout. This one is informational unless you have changed it.
Window-to-floor ratio
Checked for every habitable room — living, kitchen, bedroom, study — against 10% glazing. Two different fails hide in here:
- A habitable room with no external wall fails outright. It is landlocked, and no amount of glazing will fix it without moving it.
- A room with too little external wall for 10% is a warning with the actual percentage. Usually a wide shallow room facing the wrong way.
Fix the first by swapping that room with a neighbour that doesn't need daylight — a bathroom, a laundry, storage. Fix the second by turning the room, or by widening the opening in the editor.
Accessible bathroom clearance
The shortest internal dimension against the 1 540 mm a turning circle needs. It fails for a bathroom squeezed into leftover space, which is exactly where bathrooms get squeezed.
Fix it by taking the 200 mm from the adjacent room. It is nearly always available, and the adjacent room will not miss it.
Daylight autonomy
The share of working hours a room is usefully daylit, computed per room from its glazing and orientation.
Under 60% usually means the room is buried — too deep for its window, or facing the wrong way. The tell is a room that scores badly despite having a window: that is a proportion problem, not a glazing one, and it is solved by making the room shallower rather than by making the window bigger.
What to fix first
In order:
- Any outright code check failure. A landlocked habitable room or a 22 m egress path is a plan problem, not a detail problem, and it gets more expensive to fix the longer you leave it.
- Circulation over 15%. It is the cheapest floor area to recover, and recovering it usually improves the plan rather than compromising it.
- Daylight autonomy under 60% in a room people will actually sit in. Bedrooms can take it; a living room can't.
- Room sizes, last. They are the easiest thing to adjust in the editor and the least structural.
The order matters because the first two often fix the second two for free — a plan that loses its corridor usually gains a window.
Then lock what works
Once you have fixed something and you are happy with it, lock those rooms in the plan editor before generating again. Locked rooms come back untouched and everything else is re-solved around them, so you keep the ground you have gained instead of trading it for something new.
That is the whole loop: read the numbers, fix the worst one, lock it, go again.