A spiral staircase in an office building, providing an elegant and efficient way to move between floors

Why AI Assisted/Enabled Buildings Need Intelligent Edge Devices

by John Lannan & Margaret Naughton

We mentioned the advantages of bringing Ethernet communications to building controllers: seamless connectivity, higher bandwidth, addressability, etc. in the previous blog in this series. While building and plant controllers have been the first devices in the Building Management System (BMS) to convert from serial communications to Ethernet, going a step further and connecting edge devices assists modern BMS systems in unlocking greater insights.

Building Management System

With edge devices becoming more intelligent due to the addition of local processors and advanced sensing and control capabilities, there is more information for the BMS to take advantage of. To unlock these capabilities, the communications between the sensors and control panels also need to be upgraded.

When bringing IP communications to the edge of the building management system, devices at the edge can do more than simply be set, they can conduct 2-way conversations of status information vital to the optimal running and management of the overall building. Consider the advantages of devices at the edge performing real-time diagnostics of the building and reporting actionable insights back to the main controller to combine with other building data to make informed efficiency decisions for the entire structure. For example, real-time status and diagnostic information from Variable Air Volume (VAV) units could alert facility managers when dampers are stuck in position or not operating as expected.

A further benefit of bringing IP to the edge is the ability to perform firmware updates much faster, reducing downtime and the time required onsite by maintenance technicians.

 IP to Edge

By connecting edge devices with Ethernet, you can ensure that previously siloed systems like HVAC, occupancy, surveillance, and access control can all provide information to the central building management system to create a “single pane of glass” for all the data in the building. New patterns and efficiencies can be gleaned that previously may have been hidden when not viewed as a whole set of data points.

Modern cloud-based BMS optimization software using AI learning techniques requires significant training data to be practical. Bringing Ethernet to the edge is the most cost-effective and practical way to feed these models. With more information gathered from edge devices, this creates larger sets of control data to feed the AI engines. It can take years’ worth of data to be captured to ensure the effectiveness of the AI model, and this is where the digitization of buildings can assist in providing the necessary training data and access to ongoing data to improve and validate the AI model predictions once trained. While the first steps in building digitization have been focused on delivering a seamlessly connected network to allow centralized control of a building, the addition of intelligent edge devices is now making it possible to bring AI insights into the operation of buildings.

To better optimize building operations while reducing downtime, more information will need to be sent and retrieved from edge devices connected to the BMS. ADI has a suite of products available to make building digitalization a reality. To learn more visit analog.com/building-controllers

Find the next blog here.