Trane Technologies AI Lab
A diverse ecosystem made up of the brightest minds in AI
BrainBox AI, Trane Technologies AI Lab, is home to a multidisciplinary team of technical experts - including software engineers, data scientists, AI researchers, machine learning developers, and AI engineers. Comprised of some of the brightest minds in AI development and research, the team is primarily based in Montreal, Canada—one of the world’s leading AI hubs—with a reach and impact that extends across the globe.
Notable research partners:
Our areas of focus
Autonomous control
Development of technology using real-time data and advanced AI algorithms to enable the automated optimization of building assets using live equipment commands and control strategy assessments.
Published research papers
-
Highlighted Products
-
R&D Partnerships
Concordia University, ETS, Carleton University
Agentic AI
Development of virtual agents designed to process and contextualize large volumes of internal and external data to deliver data visualization, reasoning, and informed human-in-the-loop automated actions using advanced LLM technology, with responsible AI guardrails and mechanisms to limit hallucination.
AI predictions
Advancement of deep learning models that can anticipate building needs with strong predictive performance, supporting improved real-time control strategies.
Published research papers
-
Highlighted Products
-
R&D Partnerships
Concordia University, ETS, ASHRAE
Data infrastructure
Ongoing optimization of high-performance tech stack with real-time data processing, continuous system checks, automatic model retraining, and robust multi-layered protocols.
Published research papers
-
Highlighted Products
-
R&D Partnerships
NREL, Politecnico di Milano
Physics-informed neural networks
Combining traditional neural networks with fundamental physics principles to enhance physical interpretability and support improvements in prediction accuracy and training data needs.
Published research papers
-
Highlighted Products
-
R&D Partnerships
Polytechnique Montreal
Automated emissions reductions
Development of algorithms designed to support emissions reduction efforts, integrating real-time and forecasted emissions signals to derive consumption patterns and help prioritize cleaner energy sources.
Published research papers
-
Projects & Results
-
R&D Partnerships
University of Sydney, UC Berkley CBE, WattTime
