Crawl, Walk, Run: AI Powered Buildings

The process of maximizing the efficiency of buildings is an incremental one. Just as people must learn to crawl and then walk before they can run, building managers must first gather relevant baseline data on energy consumption associated with systems such as HVAC and lighting. As we all know, if you can’t measure it, you can’t fix it.

Next, building managers can use the data they have gathered to make incremental improvements in areas such as energy efficiency, operations, and comfort.

‘Running’, in the context of building management, takes place when data from disparate sources within a building and even across multiple buildings is integrated, analyzed, and acted upon in real time, autonomously based on AI.

In my view, the future of buildings is the following: sustainable, autonomous buildings with lower cost of maintenance.

In some ways, that future has arrived

Deploying and connecting sensors

The greatest barrier to more sustainable and profitable buildings comes from a lack of data about building systems and operations. We have to solve data gathering, both in phase one and phase two, and that is done by deploying sensors and connecting to the existing on-premises building management system.

HVAC systems consume up to 70% of the annual energy used in buildings, and variability in factors such as air flow, humidity, pressure, and temperature affect the performance of systems. Sensors that track variables such as these create a trove of information that can be exploited by AI algorithms to achieve a level of efficiency that would not otherwise be possible. Similarly, occupancy and motion sensors can generate real-time data about a building’s use that can be leveraged to enhance efficiency.

Collecting data is critical, and aggregating data from multiple sources – such as building automation and building management systems – multiplies your ability to enhance profitability and efficiency within a single building and also among several buildings in a portfolio. Because buildings are responsible for 40% of total energy usage worldwide, reducing energy consumption in this sector has global climate implications.
AI-enabled analytics can also help reduce system downtime and enhance efficiency by detecting anomalous behavior in system components such as valves and triggering a maintenance alert.

Because the maintenance and retraining of Building X algorithms occur automatically, the models continue to maintain their accuracy as conditions and their associated data change. We believe that the best AI is the one that works in the background and allows customers to focus on their business.

We have algorithms built into Building X that allow you to do some level of optimization on a dozen or so different aspects of sustainability, on different aspects of energy consumption, and on different aspects of security and operations. It's a journey, and we’ve just begun that journey.

Future opportunities with autonomous behavior

In view of the imperative to scale up smart infrastructure and automation solutions to buildings that still lack them, we have launched Siemens Xcelerator. Building X is part of this open digital business platform that enables customers to accelerate their digital transformation easier, faster, and at scale. Siemens Xcelerator is our invitation to partners, customers, and developers to collaborate – and to make AI industrial-grade. And, as more sensors are deployed and as AI becomes more powerful, we envision additional opportunities to make building management more sustainable, productive and profitable.

Today, AI can proactively detect faults before they occur, but in the future it could be configured to take actions such as triggering a work order based on the data it collects. It could also optimize space utilization to promote energy efficiency by suggesting work locations for employees within a building so that they are concentrated in a central area that is heated or cooled depending on the season.

However, as AI becomes more autonomous, it will still be guided by the parameters that building managers define for it. The algorithms could adjust the rate of air exchanges to protect the health of occupants during cold and flu seasons, for example, or a building owner might direct the system to optimize HVAC components to extend the lifespan of the system.

Phase three is autonomous behavior, where all of this happens on its own on an ongoing basis. We still have a long journey to go to add more capabilities and to optimize every single building asset. The potential benefits – for the bottom lines of organizations and for the environment we all share – are immense.

Rahul Chillar is Senior Vice President and Head of Building X at Siemens Smart Infrastructure