Cutting Through the Hype Surrounding Artificial Intelligence in Smart Buildings The talk of AI might always sound like the distant future but the technology, in different forms, is already around us.

James McHale, Managing Director, Memoori

Artificial Intelligence (AI) is already infiltrating our society and it has our smart buildings firmly in its sights; with the aim of accelerating the current wave of efficiency and insight flooding every corner of our built environment. The talk of AI might always sound like the distant future but the technology, in different forms, is already around us.

“I’ve been hearing about AI for 30 years… but it was always a future promise. What’s different now?” asks Safra A. Catz, CEO, Oracle. “First, the underlying compute capability is so much faster, meaning systems can crunch through a deluge of data almost instantaneously. Two, the ability through software to manage and analyze that data is so much better,” he answers.

AI refers to the simulation of human intelligence by machines. It enables machines to learn, test and adapt their approaches in order to emulate tasks that normally require human intuition – such as complex decision-making or visual perception. Although the idea of intelligent machines is still perceived as futuristic by many, more “simple” forms of AI are steadily moving into our daily lives.

While there is still some debate, leading experts generally divide their anticipated evolution of AI into three generations, defined by the limits of their capabilities. The technology we experience in our smartphones, tablets, automated building systems or other platforms, is restricted to Artificial Narrow Intelligence (ANI). In the future, we should expect more advanced forms when we reach Artificial General Intelligence (AGI), and then Artificial Super Intelligence (ASI). There is still a lot of benefits to be gained from ANI and, as is the nature of AI, even narrow systems learn over time to become more and more “intelligent.”

In a world awash with data, AI can help to find “the signals in the noise” by identifying anomalies and patterns, then drawing out actionable insights. AI’s advantage over human intelligence is that it can process huge volumes of data that a person or team could not feasibly analyze in a reasonable timeframe. By using historical patterns to predict future data quality outcomes, businesses can also dynamically detect anomalies that might otherwise have gone unnoticed or might only have been found much later through manual intervention.

AI’s biggest impact could be on analytics and particularly unstructured data analytics such as video analytics. “Powered by AI, video analytics can now quickly analyze enormous volumes of data contained within video streams to identify facial features, facial emotions, body shapes, and movements and gestures. Images are converted into numerical measurements (such as the distance between a person’s eyes) that combine to create a unique numerical profile of the person or object in question,” our recent report ‘Towards Data-Driven Buildings’ explains.

 

“Facial recognition systems were prohibitively expensive, required a high degree of customization, immense computing power, and expensive cameras but over time, and thanks to falling costs for computing power, advances in AI-based analytics, and the availability of high-resolution cameras at reasonable costs, services have become far more affordable,” our comprehensive report on the Market for big data software, networks, and services in buildings continues.

[an error occurred while processing this directive] Like video analytics, big data and AI are having a huge impact across smart building operations. Many of the viable emerging use cases for AI revolve around using predictive analytics to monitor the efficiency of buildings, then continually optimize based on the growing mass of data. Each of these operations is explored in further detail in the report, which brings together original analysis alongside the expertise of leading industry figures. The majority of experts seem to agree that the role of AI in smart buildings is growing rapidly.

A recent SAS study indicates that one in five companies have already implemented some form of machine learning or AI, a further 23% have tried machine learning or AI on an experimental basis. Of the remainder, 42% are exploring AI solutions but have yet to invest.

A NewVantage Partners survey indicated that investment rates are much higher amongst Fortune 1000 companies, where 97% of executives are investing in, building or launching AI based Big Data initiatives. The same executives predicted AI to be by far the most disruptive emerging technology, with 72% judging AI as “most disruptive” versus cloud computing (13%), and Blockchain (7%).

The majority of experts also agree that AI developments are still at a nascent stage, however, but they do expect a wave of deployment and adoption of AI in real-world applications in the future. Areas as diverse as speech recognition, image classification, object recognition and language maybe some of the first to develop and have a significant impact on a variety of industries. Gartner even goes so far as to say that this new general-purpose technology is just the beginning of a 75-year technology cycle that will have far-reaching implications for every industry.

“While we do believe that AI technology has the potential to be disruptive and revolutionary for many applications, it is clearly subject to a great deal of hype at the present time, and more focus needs to be dedicated towards practical use-cases rather than simply badging applications or offerings as ‘AI-driven,’” explains our Data-Driven Buildings report.