An intelligent building is one that doesn’t make its occupants look like idiots

Air conditioning controls in an office in Adelaide

I’ve spent about nearly 20 years in the post-occupancy evaluation (POE) of buildings, many of which were designed to be sustainable and low energy. Some even claimed to be intelligent buildings. If only they were. Sadly, as many working in POE will despairingly concur, unmanageable complexity is the enemy of good performance.

It’s important to remember that the term intelligent building is very much the lingua franca of the controls and building automation community. It’s not a natural phrase in architectural and engineering lexicons. You won’t find many clients using it either. It’s also a very ‘nineties’ term, like its not-so-distant relation, sick building syndrome (which, somewhat ironically, seems to have died a death). Most of the design community is now working to the ‘keep-it-simple, fabric-first’ definition of intelligence. Why? Because the high-tech approach has proved to be a mirage.

Time and time again, almost without exception, systems and technologies that rely on complex automation in order to achieve energy savings usually fail because practice doesn’t mirror design theory. Practice is a heady mix of:

• Over-complicated design with little understanding or appreciation of what occupants really want
• Design that is difficult to apply in the real world, leading to poor detailing, poor installation quality, inadequate commissioning, and the unwitting introduction of technical risks by contractual and product interfaces that go unnoticed until it’s too late
• Incompatibility of components that require constant adjustment or re-work
• Over-sensitive and/or hard to adjust controls and settings
• Excessive need for management vigilance over systems that were assumed by designers and the supply chain to be fit-and-forget, but which become fit-and-manage in practice,
• General lack of usability, compounded by false assumptions that occupants will take an interest in controlling and optimising the operation of building systems, where frankly they don’t want the responsibility
• Unexpected consequences and revenge effects: systems modulating automatically annoying occupants, systems that don’t allow enough occupant override, or which people don’t understand because the controls are not intuitive to use,
• Systems that default to an energy-saving condition rather than putting occupant expectations first (in severe cases causing a breakdown in relations between facilities managers motivated to maintain set-points come what may, exacerbated by a professional belief that things are best controlled centrally)
• The creation of a maintenance and aftercare dependency culture, where the building owner is dependent on expensive call-outs to maintain or modify the settings of digital systems for which they do not have the expertise to maintain, nor the access rights (and software) to modify themselves.

Is all this intelligent or just stupid?

The essential question a building designer needs to answer is simply this:  what problem are you trying to solve? The solution needs to be the simplest, the most appropriate, the least costly, and the most robust and reliable.  

Designers need to understand more about what end-users actually like and dislike about buildings and their systems. Although making things simple may not be the top of every designer’s list, they need to remember that buildings are intended for people – they are a means to an end not an end in themselves. Automation, in itself, should not be a goal. Building intelligence should therefore, above all else, lead to intelligible and sensible systems. Those systems shouldn’t challenge, they shouldn’t alienate, and they shouldn’t lock building owners into an expensive maintenance dependency.

Most of all, automation mustn’t disenfranchise occupants from making decisions about their working conditions, and prevent them acting upon them. It’s important to give occupants what they actually want, not what they don’t want but what designers think they ought to have.

As the author Guy Browning said: Most problems are people problems, and most people problems are communication problems. If you want to solve a communication problem, go and give someone a damn good listening to…