Buildings that respond to your every need are no longer a thing of the future. And you don’t have to work in a swanky new block to enjoy the benefits.
According to experts the majority of buildings in the UK, new or existing, will become cognitive within 10 years. These are buildings that can ‘listen’ to its internal systems and occupants to learn and maximise their own efficiency.
The Crystal building in London is thought to be one of the most sustainable buildings in the world. The jagged glass and steel building boasts more than 3,500 data points to ensure all of its systems function as effectively as possible. And it has a building energy management system which controls all electrical and mechanical systems in the building.
However such buildings are rare. And while new build might be moving, albeit quite slowly, towards cognition, the vast proportion of buildings in use today are at least five years old and as such will probably not have had the benefit of being built with this technology.
But now, as the technology becomes cheaper, companies are producing systems that can be rapidly retrofitted into existing buildings. And everyone stands to benefit: the building owners, facilities management and the tenants.
“There are a few use cases for why people do it,” says sensor manufacturer Yanzi chief executive Lars Ramfelt. “The property owner is primarily interested in energy usage, facilities management is interested in ways to better deliver its service, and the tenant wants to know how to more efficiently use the office space.”
The transformation starts with sensors. Lots of sensors.
The data is king and the analytics is King Kong.
The sensors themselves can measure temperature, humidity, CO2 and motion, but also more complex tasks requiring cameras. Most are around the size of a plug and are wireless, enabling them to be stuck onto virtually any surface anywhere, meaning that fixtures and fitting need not be taken apart to install them. The simpler ones are battery operated with around a 10 year lifespan, which is important for the uptake of this technology because as Ramfelt points out, no one would want to change the batteries on more than a thousand sensors every year.
However, the data is nothing without the analytics.
“The data is king and the analytics is King Kong,” says Ramfelt. “The person who can see the value of combining the data is going to get the most value.”
Ramfelt says most of the individual data which is collected from the sensors is straight forward, such as temperature or humidity, but the potential of the technology is unlocked when that basic data is combined during the analysis.
“The challenge is that most of the use cases are boring but when you combine them, you can make something far more interesting.”
The sensor itself acts as a small wireless computer and then uses the wireless connection to send the data to a base station which in turn sends it onto the cloud. This is where the majority of the analytics occurs.
Yanzi is currently working with IBM who carries out the analytics for the data Yanzi provides. Watson IoT (internet of things) is IBMs cognitive cloud analytics platform. This extremely powerful cognitive function means that it has the ability to learn from the data which has been collected.
One example has been sensors installed on roofs of buildings in Sweden, to measure the height of the snow that builds up. It saves time and is far safer than a employee having to climb up and measure it with a stick. But it has also had unexpected benefits.
“One user spotted an energy leak through the roof, because despite the cold weather, there was a lower depth of snow in one area where it had melted due to the escaping heat,” says Ramfelt.
Another unexpected use case involved motion sensors being installed under desks at work to allow the tenant to monitor its hot desking strategy. However, the same sensor is also now being used to analyse how much of its work force is using the desk’s ability to rise up and down, informing the next office refurbishment. In another case, it was obvious from the data that one meeting room was less popular than others. It was found that the furniture in the room was different to the others and was not fit for purpose, enabling it to be changed and the room brought into full use.
The ethics of the data which is collected is an ever present issue. However, Yanzi says it’s focusing on gathering anonymous data, rather than more detailed information about the user. For example, noise sensors can inform an occupant where it might be quietest to work, but do not recognise or record speech. Or using motion sensors instead of cameras in changing rooms; facilities are informed on how many people are using the rooms, but it does not gather or record specific information about the users.
Sending all of the information to the cloud can be extremely energy and bandwidth intensive. This is where so-called ‘edge analytics’ comes into play. Before the information is sent to the cloud or if there is no internet access, analysis of the data can be carried out at a local level, separating out the relevant and non-relevant information.
“If you want to have the next generation of power-efficient systems in terms of power and not wasting the wireless network, you need to be able to filter it and decide and cherry pick the data before sending it into the cloud,” he explains. “If you start sending everything then the battery will run out.”
He says a basic example might be where a temperature sensor might take a reading every minute. However, if the temperature remained constant then the data could instead be filtered to only notify the cloud when the state changed.
The next step is to how can you steer this kind of understanding where the end user can see the benefit directly…
“If it’s 22 degrees every minute for 16 hours, then maybe I don’t need to send it every time, maybe I just say, ’tell me when it changes significantly’,” says Ramfelt.
“By doing that you can reduce the amount of data by factors of a thousand, and things that were really expensive to install become much more cost efficient.”
Security of the data for cognitive buildings is also a very important factor in the design of the system, down to the security of each individual sensor.
“Information about the amount of hand soap being used in the bathroom may not be useful to anyone but facilities, but information on how many people are currently in the building or more crucially if you want to break into the building then it’s great to know no one is there,” says Ramfelt.
There is no doubt of the power of this technology, but ultimately Ramfelt says the technology needs to offer something tangible back to the end user for them to accept it and for it to take off. This is where the manufacturers are concentrating their efforts.
“It’s going to start happening… just like your smartphone allowed you to do lots of things five years ago. But if you don’t benefit from it, you’re not going to prioritise it.
“The next step is to how can you steer this kind of understanding where the end user can see the benefit directly and we are just at the beginning of it.”