Virtual Concentration Sensor for Energy-saving Liquid Desiccant HVAC System

Blue Frontier Climate/Computer Science/ Mathematics, 2024–25

Liaison(s): Nikhil Deshmukh, Matt Tilghman
Advisor(s): Michael Orrison
Students(s): Adia Ainsworth, Carmel Pe’er (TL), Nevaeh Thompson, Henry Trinh, Alejandro Wang

Approximately 15% of global energy consumption is used to heat and cool buildings. Blue Frontier addresses this by developing A/C technology that offers up to 90% reduction in electricity use. By using liquid desiccants, they achieve more efficient cooling and dehumidification than traditional methods. However, this requires expensive sensors, increasing upfront costs. The Clinic team is designing software-based sensors to replace the physical ones. The team is leveraging machine learning algorithms, including random forest and neural networks and physics-based models, to create these sensors, making the technology more cost-effective and scalable.