Automation isn't enough: Buildings need AI-driven intelligence

#Buildings have been among the most avid users of IoT devices. Smart buildings, in particular, use connected devices to measure everything from temperature, lighting, air quality, noise, vibration, occupancy and energy consumption, and that’s just the tip of the iceberg.

Building automation is growing in scale, with more than 6 million commercial buildings in the United States alone and an estimated 2.2 billion connected devices deployed. In 2022, the global building automation system market will reach approximately US$80 billion.

This type of automation relies on a large number of IoT devices. Many conditional action responses are automatic; if a fire is detected, the alarm is automatically triggered, often with a voice command, and the fire department is notified. This was true before the advent of the Internet of Things; now fire alarms are connected via the Internet and secondarily via cellular communications.

The value of the Internet of Things, especially in building automation, lies in two main areas:

• The data generated by indoor devices and how this data is analyzed and utilized.
• The operation and management performed by building automation systems.

Rich, continuous streams of data provide valuable insights into building operations, but there’s a problem: Large fleets of equipment generate vast amounts of data that humans alone cannot properly parse and understand. To realize the potential returns from deploying these sensors (and cameras), artificial intelligence (AI) and machine learning (ML) are needed to continuously monitor and evaluate data flows.

Automation can’t do the job alone

Until 2020, the focus of smart building systems, including building automation, was the responsibility of facilities management. Then, in addition to facilities management, the focus shifts to employee wellness and ESG programs. This opens up the need for ML-enabled capabilities.

For example, an AI system could look at air quality and find correlations with occupancy restrictions. It can also learn how to reallocate meeting rooms and cubicles related to occupancy and ventilation to maximize physical distance between employees and improve air quality, thereby reducing the chance of employees getting sick.

Artificial intelligence can also help analyze water pipe usage and water temperature to provide warnings when the risk of Legionella and other harmful pathogens increases. Legionella thrives in warm water within a specific temperature range.

The relevance of new AI capabilities does not exclude traditional functions such as tracking and managing energy consumption. With an AI-driven platform, buildings can close off areas that are not in use and try different curtain settings at different times to minimize energy use. Experiment and learn at the same time. This is a bottom-line issue that will become even more important in 2022 due to energy prices.

AI can even play a role in cleaning efficiency, identifying which tables have been used and which toilets have seen increased usage. In the age of COVID-19, facility managers are focused on cleaning.

Artificial intelligence can also greatly enhance systems that support physical security. Once the system understands what normal access and movement behavior is, it can identify abnormal behavior and raise security alerts. Other AI-driven applications can detect duress situations, abandoned objects, identify weapons, pinpoint shots—and perform emergency lockouts.

Smart infectious disease control systems can learn to utilize local infection rate data. AI systems can do things humans can’t, like stare at a wall for 20 years, looking for signs of changes in the concrete that might indicate an impending structural collapse.

Applying Artificial Intelligence to Smart Buildings

Of course, the standard starting point for a new AI-driven system is to teach it. The process begins with a data base that represents the reality the system will face. However, many will find that good basic training data for smart building systems does not exist. The answer may be to create training data by running "experiments" in physical buildings.

For example, in terms of energy consumption, you can train the system by experimentally adjusting curtains and air conditioning based on the time of day and office occupancy, thereby lowering your air conditioning bill without triggering manual overrides . Such a system could rely on temperature sensors and occupancy readings, as well as sunlight detection.

There are some basic best practices to follow. Use scientific rigor when collecting ground truth datasets, gathering data from multiple sources to increase confidence that your sample is representative.

Artificial intelligence-driven systems can learn from occupancy patterns in specific office areas and help reduce human error in space planning. Upgrading space is expensive and maintaining flexibility is critical. Space utilization and occupancy has clearly become a health issue during the pandemic. Employees may now prefer to gather on an open-air balcony or terrace to chat and drink coffee rather than in a small break room.

Where Artificial Intelligence Driven Building Management Is Going

Artificial intelligence systems can suggest changes in facility management and make building management more predictive. Speaking of responsiveness, they can also respond to unexpected challenges more effectively. A recent example; before 2020, it was impossible to identify employees who were suffering from fever and reduce the probability of infection, but this problem can be solved within current capabilities.

It takes careful consideration and investment of time to get the basic facts right. Many commercial buildings have digital twins; virtual replicas delivered by architects to building owners or managers. As a starting point, digital twins are likely to become a proving ground for AI-driven facility management and smart building management.

We expect that IT, facilities management, HR and security will become more integrated and use AI more. Joining information silos to create data flows for AI applications could have a range of benefits.

The importance of healthy workplaces, physical security and energy conservation makes it urgent to move beyond simple automation and develop reliable artificial intelligence-based building operating systems that are built on a strong foundation of up-to-date data. Any of these applications supports a strong business case; all in all, it’s a compelling argument that facilities management should look at AI-driven applications to operate smart buildings and make buildings smarter.

The above is the detailed content of Automation isn't enough: Buildings need AI-driven intelligence. For more information, please follow other related articles on the PHP Chinese website!