Lately we have seen a growth in the use of Artificial Intelligence (AI), where machines are taught to carry out human behaviours such as decision making and pattern/trend recognition. In 2016, AI companies received over $5bn in funding from investors and with the advent of home systems such as Amazon’s Echo and Google’s Alexa; it seems its widespread adoption is inevitable. However, AI does have its detractors, with Elon Musk and Stephen Hawking warning people about “a fleet of artificial intelligence-enhanced robots capable of destroying mankind”.

Regardless of whether AI turns out to be a saint or a sinner, it does have a number of potential uses within buildings. AI uses algorithms to process the data it receives and looks for patterns and trends. The more information it receives and the scenarios it experiences, the better it gets at predicting outcomes and making decisions. Rather like the Nest thermostat, which learns usage patterns within the home such as preferred settings, timings and space occupancy trends, to create its own schedule that cuts energy use and optimises the way users heat their homes.

By connecting AI to a building management system (BMS) and sensors within the asset, we can create what is in effect a ‘brain’ for the building, which is constantly processing the information it receives and responding accordingly. As it learns from the information generated by the BMS and sensors, the brain would be able to carry out tasks such as identifying incidents before they happen and improving the way activities are undertaken. This might include spotting parts that are not functioning properly and fixing them before they cause problems, or creating schedules and processes that improve the efficiency of the entire building’s daily operations. It could also react to short term changes, adjusting temperature, lighting and ventilation in response to changes in the weather, space usage and also the density of people within a room; saving money and optimising the work environment. While, in the future it may even be possible to connect brains from multiple buildings so they can share their collective learning and do things better across the board.

Despite its potential, AI is not without its flaws. As the recent NHS hack demonstrated, electronic data is vulnerable to attack and we must consider how we can secure building and other sensitive data within AI to ensure that it cannot be compromised in any way. This process must take place in conjunction with the development of these systems, so we can address the issue from the outset, rather than wait for an incident to happen and have to respond retroactively.

Although it is in its infancy, AI has tremendous potential to transform the way we operate buildings and as an industry we must start thinking now about its adoption and the associated implications. Only then can we ensure that we can reap the benefits, while managing any potential risks.

comments powered by Disqus

About the author

Andrew Pryke

Managing Director - BAM Design

Andrew leads BAM Design’s architectural, structural and interior design departments, working closely with our construction and FM divisions.

He also leads the adoption, development and integration of Building Information Modelling (BIM) adoption at BAM, to increase efficiencies at all stages of design, construction and FM.

Andrew joined BAM in 2012, following 25 years as director and project lead at James Stirling Michael Wilford & Associates and John McAslan and Partners. He has worked on projects such as The Lowry, Manchester, No 1 Poultry, London and The Royal Academy of Music, London.

A triathlete in his spare time, Andrew also applies his competitive spirit to working out the best design solutions for clients, integrating sustainable design, lean construction and full FF&E (furniture, fixtures and equipment) solutions.

“We are developing BIM as a tool for greater collaboration between design, construction and facilities management, to deliver better buildings that are easier to manage and maintain, and perform better for their users over their entire lifecycle.”

Read more articles from Andrew Pryke