BVN Usyd

We have a lovely new book – “Construction and Design Manual: Medical Practices” that goes through lots of recent (mainly European) healthcare facilities. Lots of children’s doctors, dentists, GP’s surgeries, that kind of thing, all with photographs and plans.

The beginning of the book has a section on how the plans are drawn, coloured and analysed. The most substantial example was shown to us as an example of how functional diagrams for healthcare are drawn, and how we might think about laying out our graphs.

This post is supposed to be a way of me clarifying my thoughts, as such I enthusiastically welcome comments and corrections to what I’ve written.

The plan below is of a typical ‘large multi-doctor practice’. The numbers on the rooms correspond to the numbers on the functional diagram below.

Plan of a typical 'large multi-doctor practice'

This is the same plan with the rooms coloured by function. It gives a great amount of clarity to a small image, and makes it immediately clear how the proportions of space are allocated. I think that this is a good diagram.

Coloured diagrammatic plan

They then go on to show these spaces in a more abstract relationship. The squares are sized according to their floor area, and they are connected to each other by the circulation, represented by the black lines. The small arrow indicates the connection to the outside. This gives a feeling for the proportion of differently sized rooms, and I think that it intends to display their adjacencies; I’ll outline below why I think that this isn’t the case.

"Functional diagram"

Rooms 6.4 and 1.2.2 (grey and yellow respectively) are not actually connected to the circulation in the diagram but they are connected in the plan. I’ll attribute that to user error and move on with a more general critique.

Sketch graph over the diagrammatic plan

There are a number of potentially tricky situations in this graph that should probably be pointed out to begin with. Node 34 (actually an unnumbered space in the initial diagram, but I’d imagine that it might be a waiting area) has two doors, I’ve shown this as a loop as they both connect to the circulation. The same situation occurs in node 35 (7.1 in the plan, a multi-purposive training room). There are also some slightly strange numbering situations as I missed a bit and had to go back over and fill it in later – as embarrassing as this is, it doesn’t affect the structure of the resulting graph.

Graph of connections to circulation redrawn as a loop

This diagram shows the graph that was sketched over the plan but relaxed into a loop. The loop occurs in an attempt to preserve the topology of the loop in the circulation, but could just as easily have been shown as a line.

As the example below shows, plan proximity is not a consideration in the layout of the functional diagram. Room 3.3 and 3.5 are shown to be closely connected, but are actually almost as far apart as possible in the plan.

Plan indicating rooms 3.3 & 3.5

Diagram indicating rooms 3.3 & 3.5

This seems to imply that the connecting lines aren’t actually indicative of the geometric relationships between the rooms, but more that it is possible to access one from the other. This leads me to re-draw the diagram in this format:

This readily re-inflates to the the loop graph shown above. The same graph drawn using the methodology described here in this blog would look like this:

Our methodology of drawing the adjacency graph

What this clearly show is that this is not an ideal way to visualise a plan like this either. It could be augmented with the room sizes, but it also shows no geometric relationship (although it does make the topological ones clearer).

With none of the above graph drawing methods being particularly useful for visualising a plan in this configuration we are left to try and find another more appropriate solution.

The ambiguous topology and geometry of the diagram provided by the book is unhelpful, but the presentation is excellent. I believe this to be even more harmful than it would initially seem as on a superficial reading it seems considered, and as if it contains greater insight than it really does. Having said that, I think that the overall ’look and feel’ of the book’s diagrams is a great target to chase in terms of making our own graphs look good!

The topological clarity of the conventional adjacency graph approach is great, but from all other perspectives it is pretty terrible. No useful colour and no spatial hints (either for area or proximity), make this graph almost useless to someone hoping to use it to assist with their design process. It would be interesting to see if adding the room sizes to this would be enough to make it useful, failing that, adjacency through walls could be used as a ‘weaker’ invisible edge to aid in graph organisation. (This final comment is getting very close to the Aedas Layout Tool explained on the Aedas R&D website: here)

In conclusion, I’d say that none of the above graph drawing methods are particularly satisfactory to act as as a safe aid to design. They are either unduly misleading, or just plain incomprehensible, and although the aim of this project isn’t actually produce diagrams, it could well be a very useful tangent.