BACK

Commercial HVAC Predictive Control w/ BrainBox AI

Make & share a podcast clip

Keep the exact passage and a link to the full episode. Choose up to three minutes, with enough context to be fair to the speaker.

Preview uses this episode’s audio. Browser audio export, when supported, takes the length of the selected clip. Nothing plays until you choose Preview.

Blake Standen and Omar Tabba of BrainBox AI join the HVAC School podcast to talk about commercial HVAC predictive control. This conversation focuses on applying artificial intelligence to existing building systems, including the HVAC, to predict performance under certain weather and occupancy conditions. If you have controls that can predict performance and adapt, then you can maximize energy savings.

A commercial HVAC system may include air handlers, chillers, boilers, RTUs, and all sorts of energy-consuming technologies. These systems also have controls that help direct the infrastructure, and artificial intelligence can help optimize the controls, make performance predictions based on forecast data, and make those controls communicate with foreign controls from other companies (such as via BACnet).

BrainBox AI uses a cloud to collect and hold the data it needs to predict what a building will do and help control the infrastructure. Controls react to errors, and the goal of BrainBox AI is to predict errors before they happen. For example, AI can help solve short cycling under certain weather conditions. However, buildings that use pneumatics rather than digital controls and older systems may not be good candidates for AI solutions.

One of AI’s challenges is that it requires multiple layers of training: you’re training the controls engineers, facilities staff, AND the AI itself. Another challenge of AI is that people don’t fully understand that it’s not the type of automation that takes people’s jobs; we can minimize those perceptions with education.

Blake, Omar, and Bryan also discuss:

  • Machine learning vs. artificial intelligence
  • Accuracy of prediction models
  • Virtual testing environments
  • Apathy as a challenge
  • What happens when controls go offline or are adjusted?
  • Tethered services
  • Application programming interface (API)
  • Controlling comfort and energy consumption vs. greenhouse gas emissions
  • BrainBox’s global partnership with ABB

 

Learn more at https://www.brainboxai.com/.

If you have an iPhone, subscribe to the podcast HERE, and if you have an Android phone, subscribe HERE.

Check out our handy calculators HERE.

Work through this with the trade.

Highlight a short passage, or bring a question about a specific moment into HVAC No Bull Forum.

Discuss this reference → Related forum discussions

Comments

Questionable CO Alarms w/ Bill Spohn
Same Charge, Different Numbers – Short #306
Tech, Mechanic, or Craftsman? Finding Your Fit in HVAC
Normal Subcooling, Hidden Restriction – Short #305
Airflow Serenity w/ Alex Meaney
Less Compressor, More Suction Pressure – Short #304
Mastering Your Meter Advanced Functionality with Louise Kellar (UEi Test Instruments)
Premature Coil Failure in Ductless Mini-Splits & VRF Equipment – Short #303
Advanced Refrigerant Recovery: Best Practices for Diverse Systems and Environments
The A2L & Refrigerant Transition – Short #302

Author:

To continue you need to agree to our terms.

The HVAC School site, podcast, and tech tips are
made possible by generous support from