Chengxuan Li
中文

Curriculum vitae

Chengxuan Li · PhD Researcher · AI / Software / Energy Systems, Cornell University

Short CVLong CV

Bio

1-sentence version

Chengxuan Li is a PhD candidate in Systems Engineering at Cornell University developing physics-informed and hybrid machine-learning methods for energy simulation, power-system analysis, and grid infrastructure optimization under electrification scenarios.

50 words version

Chengxuan Li is a PhD candidate in Systems Engineering at Cornell University, with a minor in ECE. His research combines machine learning, scientific computing, simulation, and computational optimization for urban energy and power systems, with applications in model calibration, smart-meter analytics, demand response, grid flexibility, and scalable digital twins for city-scale electrification and decarbonization.

100 words version

Chengxuan Li is a PhD candidate in Systems Engineering at Cornell University, with a minor in Electrical and Computer Engineering. His research combines machine learning, scientific computing, simulation, and optimization for urban energy and power systems. At Cornell’s Environmental Systems Lab, he develops data-driven and physics-informed methods for load inference, model calibration, smart-meter analytics, demand response, and urban-scale energy simulation. He is also Lead Developer of EnergyAtlas.io, a city-scale digital twin and energy modeling platform integrating building models, geospatial data, utility observations, and physical simulation to support analysis of building demand, electrification, grid flexibility, and urban decarbonization.

150 words version

Chengxuan Li is a PhD candidate in Systems Engineering at Cornell University, with a minor in Electrical and Computer Engineering. His research combines machine learning, scientific computing, simulation, and optimization for urban energy and power systems. At Cornell’s Environmental Systems Lab, he develops data-driven and physics-informed methods for load inference, model calibration, smart-meter and grid-data analytics, demand-response assessment, and scalable urban energy simulation. His work integrates physical models with statistical and machine-learning approaches to study building demand, feeder-level behavior, electrification, distributed energy resources, and grid flexibility. He is also Lead Developer of EnergyAtlas.io, a city-scale utility digital twin and energy simulation platform integrating building models, geospatial data, smart-meter observations, and physical simulation. His broader skills include AI for engineering, inverse modeling, system identification, reduced-order modeling, energy-system optimization, and practical software tools for infrastructure planning and decarbonization.

Research interests

Machine learning and AI for engineering; inverse modeling, system identification, and surrogate learning; physics-informed and physics-supervised learning; time-series modeling and load-profile inference; urban building energy modeling and reduced-order simulation; smart-meter, AMI, and grid-data analytics; demand response and load flexibility; building electrification and heat-pump adoption; distributed energy resource integration; feeder- and distribution-system modeling; scientific computing, optimization, and scalable digital twins for energy systems.

Education

  1. Aug 2024 – Present

    PhD, Systems Engineering, Minor in Electrical and Computer Engineering · Cornell University · Ithaca, NY

    • Research focused on machine learning, scientific computing, simulation, computational optimization for energy and infrastructure systems.
    • Courseworks spanning deep learning, computer vision, signal processing, Monte Carlo simulation, power systems, and applied optimization.
    • Minor in Electrical and Computer Engineering, with emphasis on power-system modeling, analysis, and grid operations.
    • Research applications include urban energy systems, smart-meter analytics, demand response, and scalable physics-based simulation.

  2. Jun 2024

    M.Arch · Architectural Association · London, UK

    • Energy engineering and benchmarking
    • Net-zero building operations and optimization
    • Post-Occupancy Evaluation (POE)

  3. Jun 2022

    BA (Hons) · Architectural Association · London, UK

    • Renewables energy systems and building energy modeling
    • Building physics and energy performance simulation
    • Environmental and sustainable buildings

Research experience

  1. Aug 2024 – Present

    PhD Researcher · Environmental Systems Lab, Cornell University · Ithaca, NY

    Research on data-driven, physics-based, and hybrid modeling methods for urban energy systems, advised by Prof. Timur Dogan, Prof. Oliver Gao, and Prof. Jacob Mays.

    • Develop machine-learning, optimization, statistical, and physics-informed methods for energy-demand inference, model calibration, and predictive modeling.
    • Research inverse modeling and surrogate-learning workflows for calibrating reduced-order and physics-based building energy models from measured data.
    • Develop scalable simulation and analysis methods for urban building energy modeling, electrification, demand response, and load flexibility.
    • Investigate time-series, system-identification, and reduced-order modeling approaches for representing building and energy-system dynamics.
    • Build computational workflows linking physical simulation, observational data, uncertainty analysis, and large-scale model evaluation.

  2. Mar 2025 – Present

    PhD Researcher, Grid Innovations · Environmental Systems Lab, Cornell University, in collaboration with Avangrid / NYSEG

    Research on data-driven modeling and flexibility assessment for electric distribution systems using utility operational and customer-load data.

    • Develop processing and analysis pipelines for smart-meter, AMI, SCADA, and related utility datasets across building, transformer, feeder, and distribution-system scales.
    • Model and predict feeder- and transformer-level load profiles, load-duration behavior, peak demand, and temporal demand patterns.
    • Develop methods for aggregating and translating customer-level demand data into distribution-grid planning and operational metrics.
    • Assess demand-response potential and grid flexibility under increasing heat-pump penetration, building electrification, and distributed energy resource integration.
    • Investigate the effects of load growth, load shifting, DER adoption, and coordinated demand response on feeder peaks, transformer loading, and distribution-system flexibility.

Professional experience

  1. Jan 2025 – Present

    Lead Developer · EnergyAtlas.io · Ithaca, NY

    Lead development of a city-scale energy digital twin and simulation platform in C#/.NET for building, utility, and urban energy analysis.

    • Design and develop the core software architecture for building-energy simulation, geospatial modeling, time-series processing, and large-scale urban analysis.
    • Develop scalable reduced-order building energy simulation and calibration workflows integrating physical models with smart-meter and other observational data.
    • Build data pipelines for building geometry, LiDAR, weather, schedules, utility data, and other heterogeneous urban datasets.
    • Develop computational methods for solar and shading analysis, geometric processing, model parameterization, and city-scale simulation.
    • Design software infrastructure for large-scale scenario analysis of electrification, energy demand, peak loads, demand response, and grid flexibility.
    • Develop APIs, Python interfaces, data models, and visualization workflows supporting reproducible analysis and integration with external research and planning tools.

  2. Jul 2023 – Sep 2023

    Environmental Engineering Consultant · Urban Systems Design MEP Engineers · London, UK

    Sustainability assessment of Google workplace properties with CBRE GWS.

    • Conduct sustainability assessment with CBRE GWS for 13+ Google workplace properties in the Americas.

Selected projects

All projects →

  1. 2026 – Present

    ShadingZip: Within-Building Selective Computation of Shading Profiles for Urban Building Energy Models

    Selective computation of representative sensors reduces ray-tracing evaluations by 67.4% at a mean WMAPE of 2.6%, while preserving full sensor and surface coverage.

  2. Jan 2025 – Present

    EnergyAtlas.io

    A city-scale utility digital twin and energy simulation platform that integrates building models, geospatial data, smart-meter observations, and physical simulation.

  3. Aug 2024 – Present

    Scalable Smart-Meter Calibration of Urban Building Energy Models

    Physics-supervised inverse surrogate learning to infer building energy model parameters from hourly smart-meter data, with uncertainty-aware calibration at urban scale.

Publications

Publications page →

  1. Chengxuan Li, Timur Dogan

    Exploration of the fidelity ladder for attic and basement representation in 5R1C Urban Building Energy Models (UBEMs)

    2026 · Submitted

  2. Chengxuan Li, Timur Dogan

    Lost in aggregation? Guidance on functional geometric fidelity in Urban Building Energy Models (UBEMs)

    2026 · Submitted

  3. Chengxuan Li, Timur Dogan

    Scalable smart-meter Urban Building Energy Model (UBEM) calibration: a physics-supervised amortized inverse surrogate learning framework

    2026 · Submitted

  4. Chengxuan Li, Timur Dogan

    Sequential coordination of precomputed Urban Building Energy Model (UBEM) responses enables scalable feeder-level flexibility assessment

    2026 · Submitted

  5. Chengxuan Li, Zihan Jimmy Wang, Timur Dogan

    ShadingZip: Within-Building Selective Computation of Shading Profiles for Urban Building Energy Models

    ASIM 2026 · Accepted

  6. Chengxuan Li, Timur Dogan

    Calendars of the City: Deterministic Schedule Libraries to Enhance UBEM Load Duration Forecasts

    IBPSA-USA SimBuild 2026

  7. Chengxuan Li, Timur Dogan

    Deriving high-fidelity residential building archetypes and typical usage patterns from national energy use surveys to enhance "initial guesses" for Urban Building Energy Model (UBEM) inputs

    Building Simulation 2025, 19th Conference of IBPSA

  8. Timur Dogan, Chengxuan Li, Hung Ming Tseng, Amber Jiayu Su, Patrick Kastner

    A Bottom-Up Urban Building Energy Model for Evaluating Thermal Load Electrification Measures

    Journal of Building Performance Simulation, 2025

    Published online July 2025; in print in volume 19, issue 2 (2026), pages 289–316.

  9. Hang Xu, Chengxuan Li, Patrick Kastner, Timur Dogan

    Understanding Urban Building Energy Consumption with Explainable Machine Learning Approaches

    CAAD Futures 2025

Presentations

  1. Jun 2026

    City-scale digital twins for building energy modeling and carbon reduction planning

    Conference talk · Cesium Developer Conference 2026, Philadelphia, PA

  2. Feb 2026

    How Cities Can Plan the Energy Transition Using Urban Building Energy Models

    Webinar · IBPSA-USA Educational Webinar

  3. Jan 2026

    Urban Decarbonization Strategies

    Webinar · IBPSA USA Simulation Showcase Webinar Series 2026

Awards

  1. May 2026

    IBPSA-USA Simulation Showcase, winner · IBPSA-USA

    $600 prize.

  2. May 2026

    Nemetschek Innovation Award, second place · Nemetschek

    €60,000 prize; second place, with no first place awarded.

  3. Apr 2026

    New York State Pollution Prevention Institute Competition, winner · New York State Pollution Prevention Institute (NYSP2I)

    $3,000 prize.

  4. Nov 2025

    NYSP2I Research Grant, Co-PI · New York State Pollution Prevention Institute (NYSP2I)

    $4,000 research grant, Co-PI with Prof. Timur Dogan.

  5. Oct 2025

    Bentley Systems Going Digital Awards 2025, Founders' Honors · Bentley Systems Going Digital Awards

  6. Oct 2025

    Holcim Building Tomorrow Scholarship, winner · Holcim

    Full sponsorship for the One Young World Summit 2025

Technical skills