Episode #2
AI Lab: The Feedback Loop Behind Trustworthy AI
How do you build AI that can be trusted in the real world? For AI to move from research into practical applications, the lab and the field need to work together. Real-world deployment creates new insights for researchers, while research and testing help strengthen the technologies being put to work. In this episode, we explore how that continuous feedback loop can make trust a foundation for responsible AI development and deployment. We hear from Foutse Khomh, Vice President of Research and Innovation, Polytechnique Montréal, about what it takes to build trustworthy AI. He explores how researchers and industry can work together to understand how AI systems behave, identify their limitations, test them in realistic conditions, and develop the safeguards and controls needed for reliable deployment. He explains why ongoing collaboration between the lab and the field is essential to continually improving AI systems. Riaz Raihan, Chief Digital Officer at Trane Technologies, takes us inside the BrainBox AI Trane Technologies AI Lab in Montreal, where researchers, engineers, product managers and building experts are developing practical AI applications for the HVAC industry. He discusses how the AI Lab connects research with real-world deployment, from autonomous building controls to new approaches to optimizing chilled-water systems, and how Trane approaches trust through security by design, compliance and partnerships. Together, they show how the ongoing exchange between research and real-world application can help build AI that is trustworthy, responsible, and capable of delivering measurable value. EPISODE CHAPTERS 00:00 Introduction - Building trust in AI 01:00 What makes AI trustworthy? 02:15 Understanding AI’s weaknesses 03:45 Testing AI for the real world 05:48 Building trust without overtrusting AI 07:45 The Swiss cheese approach to trustworthy AI 09:06 Why the lab and field need each other 12:15 Inside the BrainBox AI Trane Technologies AI Lab 15:09 From research to real-world products 16:00 Inside the AI Lab: From Aria to autonomous controls 18:45 Bringing AI to the water side of HVAC 20:18 The triangle of trust 21:51 What makes AI commercially successful? 23:57 Can AI solve the problems it creates? 27:00 From research to real-world impact LEARN MORE Trane Technologies Website [ https://www.tranetechnologies.com/ ] Scott Tew [ https://www.tranetechnologies.com/en/index/blog/author/scott-tew.html ] Dominique Silva [ https://www.tranetechnologies.com/en/index/blog/author/dominique-silva.html ] Riaz Raihan [ https://www.tranetechnologies.com/en/index/company/leadership/riaz-raihan.html ] Foutse Khomh [ https://www.polymtl.ca/expertises/en/khomh-foutse ] ABOUT HEALTHY SPACES Healthy Spaces , a podcast by Trane Technologies, brings together engineers, innovators, and industry leaders for bold conversations at the frontier of sustainable technology. In Season 6, we zero in on the moment when innovation becomes infrastructure - when breakthroughs move beyond the lab and start transforming industries. From pioneering AI research and next-generation data center cooling to climate-resilient cities and pathways to decarbonization, each episode explores the innovations at the cutting-edge of sustainable technology. Designed for leaders, builders, and anyone with a stake in the future of sustainable technology, the series offers a front-row seat to the ideas and innovations shaping what comes next. The challenges are real. The solutions are being built now.
