Design doesn’t work
Ten years ago I built a house… I designed a house and a builder…. I designed a house and a bunch of loosely coordinated subcontractors built it.
Since then I have spent a lot of time fixing things.
Most recently I’ve been fixing my downpipes and gutters as winter in the Southern Hemisphere kicks in. You’d think roof drainage was a solved problem — a mature engineering discipline — a standard solution to a standard issue. After all, we’ve been building this kind of house in this part of the world for about 200 years.
So it’s all the more puzzling when you build one, and simple things like gutters and downpipes fail to work.
The first problem is that my gutters are, for all intents and purposes, dead level. This means that water doesn’t drain well, it just sits in pools and will eventually rust out the gutter, especially if there is debris there.
The second problem I have is that all the downpipes look like this:
Yes, that 2 degree slope will probably allow water to move, but the twin right angle bends are almost purpose built to catch leaves and other debris and back-up the downpipe.
So far it hasn’t been an enormous problem but as my garden matures and my trees get bigger it’s almost inevitable. In fact I remember long hours of my childhood being stuck up a ladder, cleaning leaf litter and crap out of my parents’ gutters so that winter rains didn’t flood my bedroom.
The question that nags me, while I’m stuck up a different ladder, trying to fix all these problems, is how the hell did we get into this situation?
How, after 200 years of building houses and nearly a thousand years of engineering are we still making simple mistakes like this?
And the conclusion I have come to, is that design doesn’t work.
Design Doesn’t Work
Yes, I’m being provocative — but bear with me.
Design is about making decisions on how you are going to build something (but designers seem to indulge in making it sound like gobbledigook).
If you build the thing, then you have built the thing, not designed it.
Design is always an abstraction
If design is about making decisions on what you build and how you build it, then it is by its very nature an abstraction.
A design usually takes the form of a model, or a diagram or a specification or some other template for instructing you on how you will ultimately build the thing.
Any model must be simpler than the thing it models and so it abstracts away some detail.
For example the “National Construction Code” (NCC) of Australia quite rightly defines roof drainage in terms of performance and engineering standards and leaves the implementation up to the builder. The standard details just enough of the specification to achieve the desired outcome and no more.
The design model for my gutters and downpipes consists of the NCC standards, the engineering drawings from the builder, conventions in the buiding industry and the expert knowledge in the minds of the roof carpenters and plumbers who put the thing together.
But the standard is an average, applied over all of the houses in Australia, with a few parameters modified for local conditions. And nowhere in the plans can I find a reference to the fall of gutters or the shape of downpipes — it just isn’t there. And as for what was in the minds of the people who built it… well I’ll come to that later.
Implementation never matches design
According to the National Construction Codes, eave gutters must be designed with a minimum fall of just 1:500, a slope of about 0.2% or 0.11º.
This is what a 1:500 slope looks like:
That’s a very small amount. I’ve tried measuring the fall on my gutters and have found it impossible because the angle is so small. While I was doing this, it occurred to me that if I can’t measure it — how is the roof carpenter supposed to?
It is possible they have sophisticated measuring equipment, but I’m betting that the vast majority of them don’t — and so have assumed some rule of thumb that says “a very small angle = flat”.
This is not ncessarily the fault of the builders: the design specifies an outcome that they can’t practically achieve.
People don’t understand what is being designed for them
When our house was planned and designed, I spent innumerable hours arguing away tile selection and light switch placement with the builder.
If I knew then, what I know now, my arguments would have been quite different.
I would have fought for more insulation. I would have made them lay spare electrical conduit in every room. I would have made them build me a carport. I would have made them connect the cursed downpipes to the bloody useless soakwells. And I would have argued for gutters with tangible slopes and circular, curvaceous flowing downpipes.
In fact, I actually did argue with the builder about our gutters. I knew they were going to be a problem. I wanted a different style, that might have solved some of these problems, but ultimately lost the argument on the basis of insufficient technical knowledge (he told me it would cost more and I didn’t know any better).
People find ways to be efficient
I suspect the reason that my downpipes are shaped like the mark of Zorro has more to do with the people that built them than with any design.
Workers always find ways to be efficient.
Cutting a box downpipe into angles is much easier for a nearly 90º bend than for say a 60º bend. I know, because I’ve been slowly converting all my existing downpipes to a more graceful angle.
The definition of efficiency is the ratio of output to input. The higher the output per unit of input, the higher the efficiency.
Of course, a lot depends on your definition of “output”.
My definition in the case of gutters and downpipes, is the number of hours I spend in front of the television with a beer, instead of up a ladder cleaning gutters.
Unfortunately the roof plumber had nearly the same definition, and he was there before I was, so he wins this round.
So what’s the solution?
Don’t get me wrong – I’m not saying you shouldn’t do design.
But there’s a saying in the military: a plan never survives contact with the enemy… and a design never survives contact with reality,
Does that stop the military planning? Hell, no! They build flexible plans that can be changed and adapted on the ground to suit changing local conditions (for more on this, see my articles on Commander’s Intent and Warfighting Doctrine).
So your designs should contain outcomes and intent rather than prescribing narrow implementation detail.
But that still might not be enough.
Iterative Design isn’t
One argument I can almost hear bubbling up as I write this is that I’ve missed the point by talking about design as a once-off process. Design should be iterative... or agile… or something.
The problem I see is that if you accept my premise that design is what you do before you build a thing, then once you’ve built the thing, you’re no longer doing design.
Sure, if your design is bad enough and your problems large enough, you can “go back to the drawing board” — but that represents a wholesale restart of the process (see why the phrase exists?).
The danger is that if you incrementally modify your design as you go, you will commit the same mistakes of sub-optimisation and narrow focus as if you had not done the design at all.
What we need is something that looks like design but is significantly different.
I think what we probably need, and what I see lacking in almost every organisation, is a structured method for solving problems. Luckily such a method exists and can be found at the heart of Lean, or the Toyota Way.
The fulls scope of Lean problem solving is beyond the reach of this article, but the short version is this:
- Grasp the current state
- Identify the target/ideal state
- Examine the gap between 1 and 2, and understand why the gap exists
- Propose a set of countermeasures and perform rapid experiments to validate your thinking
- Implement successful experiments
This is not design — not as I know it.
Reconciliation
The alternative titles to this article were “The Cost of Wasted Work” and “The Value of Problem Solving” but it turned out I was writing a different article, and those two will have to wait.
I’m not bitter about the problem of the gutters.
It was inevitable that the design would fall short of my needs, and the implementation would fall short of the design.
If you go back and read the start again, you’ll see I said “I designed a house” — so I’m taking full responsibility for where we ended up. No one made more decisions about this house than me.
But luckily I have the tools to improve on the design and make the world, or at least my corner of it, a better place.
