Pratik Raj Pandey
Postdoctoral Research Associate · North Dakota State University · PhD, Mechanical Engineering
I study how buildings actually use energy, and how the people inside them shape that. My work combines field measurements, building energy simulation and machine learning to make retrofits, building models and digital twins reflect real occupant behavior and a changing climate.
Research
What I work on
Occupant behavior in simulation
Replacing fixed schedules with data-driven models of how people open windows, adjust thermostats and use their spaces, coupled directly to EnergyPlus.
Net-zero retrofits & climate resilience
Measuring what deep retrofits deliver, and how they hold up against future climate conditions and heatwaves.
Indoor air quality & ventilation
Field studies of particulate matter, VOCs and ventilation in homes and residence halls, and how building systems affect them.
Calibration & digital twins
Calibrating building energy models against sub-metered data, and building digital twins of HVAC and electrical systems in EnergyPlus and Modelica.
Selected projects
Projects
2021 – 2025
Two years of data from a dormitory going net-zero
Led data collection, analysis and energy modeling for a NYSERDA-funded deep energy retrofit that turned 1970s student dormitories into high-performance, net-zero buildings. The study tracked indoor environment, occupant behavior and disaggregated energy use in 16 units before and after the retrofit, and the dataset is now openly published in Scientific Data.
2021 – 2025
Teaching building models how people actually use windows
Trained deep neural networks (over 96% accuracy) to predict window-opening behavior and coupled them to EnergyPlus through its Python API in place of static schedules. The results showed that conventional simulations can underestimate energy lost through windows by more than 580 kWh per unit each year. Calibrated models with genetic algorithms to under 15% error against measured data, and extended the approach with PINN and GAN models and a meta-model applied across 5.1 million New York State building models.
2026 – present
High-fidelity digital twins of building systems
Developing, validating and deploying digital twin models of building electrical and HVAC systems, using EnergyPlus and Modelica to analyze and optimize building energy performance.
2019 – 2021
Which building systems clear indoor particulate matter?
Collected field data in Kansas homes to compare how windows, central HVAC and kitchen range hoods remove fine particulate matter (PM2.5), and built a model to predict concentrations from how long each system runs. Opening windows was fastest; range hoods often had little effect in practice because they were run too briefly. Also examined how COVID-19 lockdowns changed occupant behavior and indoor air quality.
2020
Safer masks for students who rely on lip-reading
Designed a test rig with real-time particle monitors to measure the filtration of eight mask types, including cloth masks with transparent windows used in speech and hearing education. Multi-layer windowed cloth masks filtered more than 70% of PM1, PM2.5 and PM10, which is comparable to surgical masks and better than standard double-layer cotton masks.
Publications
Papers & talks
Most recent first. The full list with citations is on Google Scholar.
Journal articles
Conference papers
Reports & theses
Background
Experience & education
- 2026 –Postdoctoral Research Associate
North Dakota State University, Fargo ND Digital twins of building HVAC and electrical systems; EnergyPlus and Modelica simulation. - 2026 –Academic Research Specialist
Mercor (remote) Evaluating AI model responses on HVAC design, thermodynamics and building energy against engineering principles and ASHRAE standards. - 2021 – 2026Graduate Research Assistant
BEST Laboratory, Syracuse University Net-zero retrofit living lab, occupant behavior modeling, building sensor networks, IEA Annex 79 reports; mentored summer interns in energy modeling and Python. - 2019 – 2021Graduate Research Assistant
Sustainable Building Energy & Environment Lab, Wichita State University Residential indoor air quality, COVID-19 occupant behavior, mask filtration testing.
- 2025PhD, Mechanical Engineering
Syracuse University - 2021MS, Mechanical Engineering
Wichita State University - BS, Mechanical Engineering
Wichita State University - Tools
EnergyPlus, Modelica, OpenStudio, DesignBuilder, CONTAM, Ansys Fluent · Python (pandas, PyTorch, eppy, PyGAD), R, SQL · Tableau, ArcGIS Pro - Professional societies
ASHRAE, IAQA
Contact
Let's talk
I'm always glad to hear from researchers, practitioners and students working on building energy, retrofits or indoor environments.
[email protected] · LinkedIn · Fargo, North Dakota