AUTOMATIC FOR THE PEOPLE
As you stand at baggage reclaim you open an app on your phone. This tells you that the house is currently at seventeen degrees, the heating is on, and that it will hit your preferred temperature in an hour just as you turn the key in the lock. Gone are the days of returning to a cold house with the thermostat having been left in ‘holiday’ mode.
This is one feature of a new generation of home thermostats, the most well-known of which is designed by a subsidiary of one of the large Californian technology companies. These thermostats learn about you. They record your daily routine and your preferred temperatures. They also learn about your house and its heating system. They can predict when to turn on the heating system, based on your preferences and the forecasted external conditions.
This intelligent behaviour is being calculated using a Machine Learning algorithm. Machine Learning is a subset of Artificial Intelligence (AI) where algorithms are designed to learn and improve from experience rather than being explicitly programmed, as in traditional systems.
In order to learn, large amounts of data are required to train the algorithm. It is no coincidence that the current surge in interest in AI follows from the huge datasets which have been generated through the world wide web and smartphones. Learning thermostats are effectively Internet of Things (IoT) enabled devices. So-called because, they are able to use the internet to upload data from sensors to web-based servers. The collected data is then used to refine operational behaviour.
What could be the potential of this idea if scaled up to larger buildings and systems? A well publicised example is one of the world’s largest digital companies employing machine learning algorithms to reduce the energy associated with data centre cooling by 40%, leading to an overall energy saving of 15%.
The key reasons why these data centres were good candidates for AI technologies included:
An abundance of training data existed from historical monitoring from a multitude of sensors
The systems which control the internal climate are complex and there are thousands of ways which individual items of equipment can interact with each other
This means that for every possible permutation of internal server load and external weather data there will be an optimum operating scenario of the various components
These optimised scenarios cannot be arrived at by human intuition alone
The savings achieved were startling given that these facilities were already highly engineered, given the significant energy costs as well as the aspiration the company has for its data centres to be zero-carbon.
Machine learning as a tool essentially looks for patterns in training data to develop an algorithm which produces a reliable output from given inputs. Once the learning is complete, the algorithm can be used as a prediction tool for optimising outcomes.
Initially when used to optimise the performance of data centres, algorithms were used to provide recommendations for human operators to implement. This was to ensure the system did not try to move to a state which would have put the datacentre at risk. At the end of last year, it was announced the algorithms had been allowed to make changes to the control system directly, within predefined safe ranges for the equipment.
This direct control of the datacentre by a machine learning based AI agent is exciting for the future of buildings and the built environment as we seek to decarbonise. Non-domestic buildings are generally unique and involve a complex arrangement of services, ranging from chilled water plant through to automated openable windows. Control systems for buildings are in turn unique and programmed by hand to predicted operating scenarios developed by the designer which is inherently limiting as with the operation of data centres.
The key to advancing building control systems towards an AI enabled future is the adoption of communication platforms which can support its deployment. Currently many companies offer data analytics tools which continuously read the status of points in a building management system. These can provide insights and allow building managers to make informed decisions. However, the modification of the architecture in a building control system requires human intervention. Moving to a point where it’s possible to ‘write’ to the BMS in a meaningful way may take some time.
In the meantime, when designing buildings, it’s important to carefully consider the platform for communication-based systems, their interoperability, the ability to collect data, and the security of that data. Bear in mind that in the future, data from an access control system could be as useful to defining the operation of building systems as a temperature sensor. In operation, look after your data, keep BIM models up to date, store data in accordance with approved standards and consider the use of a digital twin.
This article was originally published in CIBSE Journal, April 2019