Webinar

Mitsubishi Electric – Building High-Fidelity HVAC Models with Dyad and Scientific Machine Learning

Webinar

Mitsubishi Electric – Building High-Fidelity HVAC Models with Dyad and Scientific Machine Learning

Event Date & Time

EDT

Speakers

Share

Event Date & Time

EDT

Speakers

Share

Join Christopher Laughman (Mitsubishi Electric Research Laboratories) and Avinash Subramanian (JuliaHub) for a webinar exploring how advanced simulation tools can transform HVAC system modeling.

Accurately estimating refrigerant mass in vapor-compression systems such as air conditioners and heat pumps, is essential for both performance and environmental assessment. Traditional methods are invasive, costly, or insufficiently precise.

In this webinar, Laughman and Subramanian will share how MERL used Dyad’s ModelingToolkit-based workflows together with machine learning techniques to develop a non-invasive state estimation approach. By leveraging pressure and temperature data, their method can predict refrigerant mass and other hard-to-measure system variables with high accuracy, achieving errors of less than 2%.

What you’ll learn:

  • The limitations of traditional diagnostic methods in HVAC systems

  • How Dyad enables high-fidelity modeling of vapor-compression cycles

  • Combining physics-based models and machine learning for improved state estimation

  • A MERL case study demonstrating <2% error in refrigerant mass prediction

  • The potential of simulation-driven approaches to improve diagnostics, efficiency, and sustainability in HVAC

Click here to sign-up for the webinar.

Speakers

Christopher R. Laughman received the S.B. and M.Eng. degrees in electrical engineering and computer science and the Ph.D. degree in architecture from the Massachusetts Institute of Technology, Cambridge, MA, USA, in 1999, 2001, and 2008, respectively. He has been with Mitsubishi Electric Research Laboratories, Cambridge, since 2008, where he currently holds the position of a Senior Principal Research Scientist and is the Senior Team Leader of the Multiphysical Systems Team. His research interests include the modeling, simulation, control, and optimization of large-scale multiphysical systems, with an emphasis on multiphase thermofluid applications.

Avinash Subramanian is a Software Engineer at JuliaHub, specializing in simulation, control, and optimization. He leads the development of Dyad HVAC and thermal-fluid system models, enabling high-speed simulation of complex two-phase flows through advanced numerical methods. He also contributes to JuliaHub’s Optimal Uncertainty Quantification efforts, building algorithms for distributionally robust optimization in safety-critical systems. Prior to JuliaHub, Avinash was a graduate researcher at MIT’s Process Systems Engineering Laboratory and a lecturer at NTNU, Norway, where he taught Mixed-Integer and Nonconvex Optimization. He holds a PhD in Energy and Process Engineering from the Norwegian University of Science and Technology (NTNU).

Speakers

Christopher R. Laughman received the S.B. and M.Eng. degrees in electrical engineering and computer science and the Ph.D. degree in architecture from the Massachusetts Institute of Technology, Cambridge, MA, USA, in 1999, 2001, and 2008, respectively. He has been with Mitsubishi Electric Research Laboratories, Cambridge, since 2008, where he currently holds the position of a Senior Principal Research Scientist and is the Senior Team Leader of the Multiphysical Systems Team. His research interests include the modeling, simulation, control, and optimization of large-scale multiphysical systems, with an emphasis on multiphase thermofluid applications.

Avinash Subramanian is a Software Engineer at JuliaHub, specializing in simulation, control, and optimization. He leads the development of Dyad HVAC and thermal-fluid system models, enabling high-speed simulation of complex two-phase flows through advanced numerical methods. He also contributes to JuliaHub’s Optimal Uncertainty Quantification efforts, building algorithms for distributionally robust optimization in safety-critical systems. Prior to JuliaHub, Avinash was a graduate researcher at MIT’s Process Systems Engineering Laboratory and a lecturer at NTNU, Norway, where he taught Mixed-Integer and Nonconvex Optimization. He holds a PhD in Energy and Process Engineering from the Norwegian University of Science and Technology (NTNU).

Speakers

Christopher R. Laughman received the S.B. and M.Eng. degrees in electrical engineering and computer science and the Ph.D. degree in architecture from the Massachusetts Institute of Technology, Cambridge, MA, USA, in 1999, 2001, and 2008, respectively. He has been with Mitsubishi Electric Research Laboratories, Cambridge, since 2008, where he currently holds the position of a Senior Principal Research Scientist and is the Senior Team Leader of the Multiphysical Systems Team. His research interests include the modeling, simulation, control, and optimization of large-scale multiphysical systems, with an emphasis on multiphase thermofluid applications.

Avinash Subramanian is a Software Engineer at JuliaHub, specializing in simulation, control, and optimization. He leads the development of Dyad HVAC and thermal-fluid system models, enabling high-speed simulation of complex two-phase flows through advanced numerical methods. He also contributes to JuliaHub’s Optimal Uncertainty Quantification efforts, building algorithms for distributionally robust optimization in safety-critical systems. Prior to JuliaHub, Avinash was a graduate researcher at MIT’s Process Systems Engineering Laboratory and a lecturer at NTNU, Norway, where he taught Mixed-Integer and Nonconvex Optimization. He holds a PhD in Energy and Process Engineering from the Norwegian University of Science and Technology (NTNU).

/

/

Mitsubishi Electric – Building High-Fidelity HVAC Models with Dyad and Scientific Machine Learning