Page 63 - RC21 EDGE Summer Issue

AI AND CLOUD COMPUTING – Continued from page 51 power, has made it possible to integrate control platforms
without the need for the open protocol hardware devices.”
The NFL’s NEXT GEN STATS is a great example of harnessing the power of AI and cloud computing. With sensors attached to players’ shoulder pads and the game ball, movements of the player and the ball are charted within inches. The data is shared with the television broadcasters and at the stadium, giving the NFL data
on all players and providing an incredibly enhanced fan experience.
BrainBox AI, whose deep learning energy-saving AI was recently recognized as a TIME Best Inventions in 2020, can not only integrate via BACNET protocols and Niagara drivers, but also via a cloud-to-cloud connection. “Some large real estate and OEM companies have started to
put BMS and other control data into the cloud in a way that allows for two-way read/write communication. With this connection, no hardware is installed on-site as the connection is made in the cloud,” said Omar Tabba, Brain Box AI’s vice president, Product and Solutions.
In some cases, even legacy systems without internet connectivity can be integrated. With the help of AI, software can be written and dropped into gateways that connect to the legacy system via its local comm port.
This allows the data to be extracted and uploaded in the same fashion as modern systems, allowing more buildings to benefit from smart building technology.
Once the access to all the data points is achieved, the database creation can begin.
ThoughtWire, a company that specializes in digital twins for buildings, uses the healthcare system as
a model to their database creation. As Jason O’Neil, director, Product and Data, Smart Buildings, explained, “Healthcare has a common interoperability language where vendors align to a common protocol to share patient information from different information platforms so that doctors can see all information when studying
a patient’s history. That doesn’t exist in smart buildings. So, to bridge the gap, we tag all the data points from the various platforms with a common data structure.”
This common data structure is critical as it allows the analytics to understand what is being analyzed and make important conclusions without discrete sensors. For example, patterns in conference room scheduling may lead to the conclusion that a building will be unoccupied for a certain period that is not otherwise a holiday. This information can be taken advantage of by adjusting setpoints, turning off lights, etc.
63