Zen and the art of building maintenance

I am in the throws of reading Zen and the Art of Motorcycle Maintenance by Robert Persig. In my view it is a philosophical book that challenges the reader to find beauty in technology, in maths and in reason. I am no book reviewer, so I will only add that Persig uses the motorcycle as an analogy for different ways of thinking. In particular, he talks about the idea of an ‘intellectual knife’: the tool with which we slice up a problem into its components before we set about resolving it. So each evening on the metro, I have been reading about different ways a motorcycle can be divided up in order to explain its different functions, and it has struck me how similar this exercise is to what I am doing at work in order to calculate the cost of the cost of a skyscraper.

One way to add up the cost is to start at the top left-hand corner, and work your way down to the bottom right-hand corner, counting up all the lumps of concrete and steel along the way. This method would be ideal for a bungalow but not for a high-rise as for starters it fails to take into the repetition in the structure. So the first use of the knife is to cut the building up into repeating chunks. Now in the case of this building, like the Gherkin, this building is curvy, so no two floors are identical. The knife is therefore used to cut the building up into chunks whose dimensions are broadly similar, so that mean values for these chunks can be used.

Am I boring you yet? Then look at the photo. I put it in to spruce up what on the surface might otherwise be an apparently boring entry. It’s the view from my office (prizes for anyone who can spot the Eiffel tower). Refreshed? Right, lets carry on…

So we have our broadly similar chunks of building: can’t we start counting? Well yes, but if you want to automate the process you have to put into Excel. I do want to automate it because this project is constantly changing and I want to quickly be able to modify the calculation. This is really where the headaches begin. So often have I marched into writing an Excel spreadsheet only to find that when I am waste-deep in it, it becomes very complicated, difficult to verify and impossible for anyone else to follow. This happened to me last week on this same project. I spent the weekend thinking that there must be a better way. Before I started again I set out the main things I wanted to achieve when doing the calculation again. It has to be easy to enter the data, easy to modify the data, easy to verify the results and easy for somebody else to follow.

If there are any readers left, I want to illustrate the problems that these objectives can cause. I won’t go into how I solved them because whereas the objectives are general and can be applied to the automation of other engineering problems, the solution is specific, of less ‘interest’ to others, and I have that recorded in the form of the spreadsheet itself.

Starting with entering the data, it is very well to count up all the similar columns but how can you be certain that they have all been counted. One answer is to create a big grid with all the stories, to cut and paste in all the similar elements and then to count them all up. Whilst this approach starts off very pleasing to look at, (I think this approach uses Excel well) it quickly becomes unwieldy. In order to simplify things, it is necessary to only put in the absolute minimum of information, putting the rest of the information perhaps on another sheet. The risk here is that the sheet quickly becomes difficult for someone else to follow. The other problem with hiding information elsewhere is that it also becomes difficult to modify quickly.

One might conclude that ease of modification is at the cost of simplicity. However with careful application of the intellectual knife, I don’t necessarily think that this is so. Experience of this sort of calculation, something that I don’t have much of, would give an idea of which variables are more likely to vary and so which ones should be easier to change. For example it may be that the thickness of the core walls are much less likely to change than the thickness of the floor.

Moving on from the depths of dullness and back towards a level of interest that might only correspond to vaguely dull, the one thing I haven’t talked about is how to check you got the right answer, because after all, that’s all that matters. Verifying my procedure and checking that it gives me a reasonable answer has taken me so long (most of the week in fact) that I wonder if using a computer has saved me any time (remember citizens that that is what computers are for…). Trying to unravel what a string of cell names in a formula actually means is the bane of my nascent working life! It doesn’t help that I have no sense of what the answer should be.

That all changed yesterday when I spent the afternoon looking at the final cost add-up of another building designed by this company. By looking at what the price of the foundations was as a percentage of the total cost, say, I instantly had a ball park figure to head for. I then used this rubric to look for where my answers were way off par. Sure enough, where there were discrepancies, there were mistakes.

It makes me wonder why bother doing any of this calculation afresh. The two match up so well that surely one could take the price of the first, modify for inflation, add a bit, and be done.

To conclude, after a week of work on this calculation, I am happy with the result. I have gone into some detail about what I have done because for one thing, a great deal of reflection would quickly have been forgotten as soon as I move onto the next thing. I fear the hours that can be spent in front of computer screen with nothing to show for it. I hope that at the very least I will be able to apply what I have learnt here to the next problem, and more ambitiously that the fruits of my labour will be a rethink of the way that these problems are tackled, which strikes me, albeit as a novice, as inefficient.

It seems entirely appropriate that ‘Zen and the Art of Motorcycle Maintenance’ has inspired me to think in depth about what I am doing. It is unfortunate for me however that motorcycles sound a lot more sexy than volumes of concrete and Excel spreadsheets. Thanks for reading.