If smart buildings are to deliver the energy savings that are hoped for they will need to evolve from their present reactive state and become truly predictive.
When you consider the data, it is obvious why energy-efficient or smart buildings are very high on the current agenda: worldwide, buildings consume some 42 per cent of all electricity generated and by 2025 they will be the largest emitters of greenhouse gases on the planet.
'Smart building' is a term that is very popular among opinion-formers, but what exactly is it, and how can it be employed to reduce energy use?
"I think the first thing to establish is what the majority of people think a smart building is," Simon Robinson, technical director, WSP Group, says. "It may seem like a fundamental question. However, some think it simply means automated or high-tech or intelligent; others think it means low environmental impact, and then there are others who think it means a combination of these.
"For me, smart buildings imply those [buildings] that look after themselves against changing parameters rather than set parameters."
"Engineering philosophies are quite often driven by different agendas," says Cameron Steel CEng, IET Sector chair - Built Environment. "A sharp sales engineer may sell you a piece of technology to save energy for the end user, but without consideration for the embodied energy in that piece of technology. Someone else will perhaps look at the bigger picture and have a different view."
Automated buildings have been around since the 1960s and have increased in complexity since then. These buildings include such technology as optimisation and compensation of heating systems, where the control system looks at both external and internal temperatures over 24 hours before deciding on switch-on times. Maybe lighting that turns itself on when it gets dark is a simple automation to achieve, as are doors that open when someone approaches - but both have been around since before the 1970s. Fire alarm systems that monitor their condition and can track the path of a fire from its source have been around since the 1980s.
There is an argument among those in the built environment as to whether the building intelligence or the intelligence of the occupants is of the greater importance. Most believe that it is a combination of intelligent automation that is backed up by educated users and visible prompts. "I'm sure that user behaviour will adapt but not necessarily to the technology," Steel says. "We each have our own requirements for the space around us and will strive to make it comfortable. That may be a potted plant on the desk, or a portable electric heater hidden from the facilities manager because we feel too cold, regardless of what the building considers to be adequate.
"The last two decades has seen smart technologies that are little more than gimmicks. Maybe it's just my observation, but numerous ways to close your curtains or blinds appears to be some kind of benchmark. All of the above examples are simple control systems and are reactive. The future for me is proactive or even predictive control."
An example could be people tracking via access passes allowing building services control systems to build knowledge of when and where people gather and alter heating, cooling, lighting and ventilation accordingly.
"I think we need more visibility in intelligent building control systems because people engage more if they know what is happening and why," Robinson adds. "Very often they are complex systems that sit in the background. I don't know if there is any specific evidence. However, from my observations it seems that people alter their behaviour if they know that behaviour is having an effect on the energy they are using.
"Although not related to buildings, one good example of this is the data available to the driver in a car. Most people I know tend to drive more economically if they set the display in their car to show miles per gallon being achieved. Similar systems in buildings could display real-time energy usage and also let people know when things are happening."
One example of this could be the lighting installation in large open-plan offices. Simple lighting control can turn off the row of lights adjacent to a window if the light level rises above a pre-set threshold. This can cause people to complain, even though there is sufficient illumination from daylight.
Solutions tend to be based around dimming the lighting rather than turning it off. However, a display system in offices could inform staff that the illumination levels outside had reached a particular level and the lighting adjacent to the window was being switched off.
Another example could relate to mixed-mode ventilation systems where a combination of natural and mechanical vent is used. If people feel a room is stuffy in winter and the mechanical vent is not working, they will complain. It may be, however, that the CO2 sensor hasn't recorded a CO2 level above the threshold to activate the ventilation so there would be no need for it to operate.
Again, a visual display showing the actual CO2 parts per million content in the room, along with atmospheric content and safe upper levels for reference, would better inform and educate the people who have to live with the system.
"I did visit Siemens in its Zug factory six or seven years ago and it was looking then at integrated systems that coupled access control with addressable lighting control and associated environmental controls too," Steel says. "It was also integrating the fire alarm systems with emergency lighting too. Finding a way, in larger buildings in emergency situations, of avoiding the seat of a fire is the next obvious step - perhaps it is already being done.
"It comes down to cost. If you have a large building with a transient population then you will need tight control and a 'big brother' system, because turning the lights off on the way out is someone else's problem. In a smaller building, passive approaches and education of the users is perhaps more appropriate, cheaper and more applicable."
According to the dictionary, 'smart' means being "capable of making adjustments that resemble human decisions, especially in response to changing circumstances". "The dictionary definition is interesting," Robinson says. "If we applied this definition, I don't think we could claim any building is really smart.
"Only this morning I was looking at a radiator at home caught in the sunlight. There were deep shadows across it giving it a number of shades of grey as well as the white which it is painted. If we asked even the most sophisticated computer what colour the radiator was, assuming it has the same viewing angle I did, I doubt it would come up with white as an answer. This simple observational task is beyond cold analysis of data and requires an application of prior knowledge and assumption. It was white yesterday and therefore in all likelihood it is white today, therefore you could say I guessed.
"If you apply the argument to human decisions, in particular related to buildings, I turn up the heating when I feel cold. Today I might feel cold when the ambient temperature is 20°C and tomorrow I might feel cold when it's 21°C. Control at this level cannot be automated and would require the input of the user (me) to make final adjustment.
"This is only based on the parameters of achieving a comfortable temperature. If we add in the requirement to reduce the amount of energy used and hence reduce directly related carbon emissions, a number of factors have to be considered, which means the buildings users and their interaction are inextricably linked with the smart concept."
As pressure continues to grow for buildings to become more energy efficient, the demand for automation will continue to grow. "It strikes me that it is the next stage of micro control, with such feedback and knowledge, when the controls understand what the occupiers want and the control system starts to understand that one person likes the temperature up a degree on one floor and another prefers it slightly cooler," Jerry Whitlock, technical director and head of central technical services at hurleypalmerflatt, says. "This could link to access control, for instance, as the system learns about the occupiers movements, preferences and habits."
A slightly scary big brother outlook. *