While electric vehicles and aerospace capture public attention in the green energy transition, the silent giant of global carbon emissions is sitting right above our heads. Buildings account for roughly 40% of all domestic energy consumption in the US—a baseline that is scaling dramatically across the globe as shifting climates drive intense new demands for air conditioning and cooling.
In this interview, engineering expert Merle Laughman breaks down why solving the HVAC efficiency problem requires moving past unanchored "big data" machine learning. Vapor compression cycles are highly dimensional systems that suffer heavily from the curse of dimensionality. Crucially, the most vital operational windows occur on the hottest and coldest days of the year—leaving systems running in highly dangerous, data-starved corner cases where nominal ML algorithms fail.
Discover how Dyad and Julia provide the ultimate toolkit for declarative, physics-based modeling. Learn about Merle’s pioneering research into high-performance, patch-based B-spline thermodynamic property models—which run thousands of times faster than legacy lookups—and see how bridging imperative programming with declarative simulation environments enables smart control horizons that leverage solar grids, building thermal mass, and predictive pre-cooling.
