- Bluon has launched a generative AI tool for HVAC technicians and is making a beta version of it available free for public use and feedback through Oct. 31 .
- The chat interface, MasterMechanic, aims to help HVAC operators identify, pinpoint and address complex and specific HVAC-related problems using only natural language, the software company said.
- The MasterMechanic rollout comes at a time of a growing need for facilities managers to access specialized language or knowledge, especially amid a labor shortage in skilled trades.
The need for well-functioning HVAC units has become increasingly critical against the backdrop of continued concerns about indoor air quality. Following updated building ventilation guidelines from the Centers for Disease Control and Prevention earlier this year, the American Society of Heating, Refrigerating and Air-Conditioning Engineers published its first guidelines on air ventilation and quality in June. Standard 241P, Control of Infectious Aerosols, is aimed at establishing minimum HVAC requirements for commercial, hospitality and healthcare spaces. Municipalities commonly adopt ASHRAE’s standards as part of their building codes.
Still, constrained budgets and rising temperatures are exacerbating HVAC challenges for facilities managers in environments like schools, many of which have deteriorating HVAC systems. A 2020 U.S. Government Accountability Office study found that more than 36,000 U.S. schools need to replace or update their HVAC systems. Recently, nearly 400 maintenance tickets were submitted to request air conditioning repair in Tucson schools within a two-week period, while school districts across at least six states were forced to either close or dismiss students early due to broken or non-existing cooling systems.
Bluon, based in Irvine, California, said MasterMechanic is designed to provide step-by-step answers to questions HVAC operators may have during upgrades, replacements, repairs and troubleshooting activities. The assistive application is trained on Bluon’s body of HVAC data, which includes 50,000 real-world problem-solution scenarios as well as insights from “some of the top mechanics in the trade,” the firm said in a statement, adding that the system understands a breadth of HVAC issues in many types of systems, ranging from complex commercial structures to residential buildings.
The machine-learning approaches MasterMechanic uses include reinforcement learning with human feedback, large language models and retrieval augmented generation in high dimensional vector space, Bluon Senior Vice President Adam Curry said in the release. “We've been really impressed with the progress in the machine learning community; it's helped us maximize the possibilities of Bluon's immense set of HVAC data,” he stated.
Peter Capuciati, chairman and CEO of Bluon, added that the firm has an ambitious road map for the product and intends to continue expanding its capabilities.