Transportation Electrification
Where EV charging is needed, who invests in it, and how fleets make the switch.
Why it matters
Transportation is the second-largest energy consumer in the U.S., behind electric power generation, and about 93% of the energy it uses comes from petroleum. Electrification cuts emissions only if charging is available where people live and work, and if fleets can see that zero-emission trucks perform in real service.
What we do
We model how EV drivers decide when and where to charge, and we use those models to plan charging infrastructure that matches demand instead of guessing at it. We are also integrating EV charging behavior into column-generation-based traffic assignment, so network models capture how charging changes traffic.
Our projects put the models to work. In Kansas City we evaluated streetlights as low-cost curbside chargers for residents without home charging; the work drew national and international media coverage in 2025. In Pennsylvania we provide technical assistance to the Department of Environmental Protection on replacing Class 4–8 diesel trucks with zero-emission vehicles, including data collection and analysis of how those trucks perform in real fleet applications.
What we work on
Charging demand and driver behavior
When and where drivers charge, and why: range anxiety learned from the initial state of charge, charging behavior patterns from functional data analysis, and multimodal large language models that read satellite imagery to predict charging demand.
- VisionEV: Multimodal Large Language Models for Spatially Aware Electric Vehicle Charging Demand Prediction Using Satellite Imagery (TR Part D, 2026)
- Learning Electric Vehicle Driver Range Anxiety with an Initial State of Charge-Oriented Gradient Boosting Approach (JITS, 2023)
- Analysis of Electric Vehicle Charging Behavior Patterns with Function Principal Component Analysis Approach (JAT, 2020)
Charging infrastructure planning and investment
Demand–supply coupled models that site charging stations while accounting for how demand responds, and a unified framework for deciding the mix of public and private investment.
- Learning-Based Demand-Supply-Coupled Charging Station Location Problem for Electric Vehicle Demand Management (TR Part D, 2023)
- Public or Private? A Unified Demand–Supply Coupled Framework for EV Infrastructure Investment (TR Part D, 2026)
Streetlight charging
With the U.S. Department of Energy and the City of Kansas City, we studied retrofitting existing streetlights as curbside chargers for residents of apartments and multi-unit dwellings who cannot charge at home.
- Equitable Urban Electric Vehicle Charging: Feasibility and Benefits of Streetlight Charging in Kansas City Right-of-Way (ASCE J. Urban Plan. Dev., 2025)
- Kansas City, Missouri, Streetlight Electric Vehicle Charging: Strategies and Challenges for Site Selection of Streetlight Electric Vehicle Infrastructure in Kansas City, Missouri (DOE report, 2024)
Medium- and heavy-duty fleet electrification
Technical support for Pennsylvania DEP's zero-emission truck programs, electrification of terminal trucks in freight yards, Penn State's own zero-emission fleet, and DOE-funded training for emergency responders who work with zero-emission trucks.
Representative work
VisionEV: Multimodal Large Language Models for Spatially Aware Electric Vehicle Charging Demand Prediction Using Satellite Imagery
Uses multimodal large language models on satellite imagery to predict where EV charging demand will arise, adding spatial context that tabular models miss.
Public or Private? A Unified Demand–Supply Coupled Framework for EV Infrastructure Investment
A single demand–supply coupled framework for deciding how public and private investment in EV charging infrastructure should be combined.
Equitable Urban Electric Vehicle Charging: Feasibility and Benefits of Streetlight Charging in Kansas City Right-of-Way
Evaluates the feasibility and benefits of converting streetlights in the Kansas City right-of-way into EV chargers for residents without home charging.
Learning-Based Demand-Supply-Coupled Charging Station Location Problem for Electric Vehicle Demand Management
Couples learning-based EV charging demand prediction with the charging station location problem, so siting decisions and demand management are planned together.
Sponsored projects
All projectsDriving PA Forward Program Technical Assistance
Pennsylvania Department of Environmental Protection
2026–2031
PIImmersive and Interactive Training on Medium and Heavy-Duty Zero-Emission Vehicles and Infrastructure Technologies for Critical Emergency Response Workers
U.S. Department of Energy
2025–2028
PIPennsylvania Medium and Heavy-Duty Zero-Emission Vehicle Program Data Collection and Analysis (Phase 2: Post Award)
Pennsylvania Department of Environmental Protection
2024–2026
PIPenn State Zero-Emission Fleet Implementation
Penn State Student Fee Board
2023–2025
Co-PIPennsylvania Medium and Heavy-Duty Zero-Emission Vehicle Program Data Collection and Analysis (Phase 1: Pre-Application Planning)
Pennsylvania Department of Environmental Protection
2022–2023
PIElectrifying Terminal Trucks in Unincentivized Markets
U.S. Department of Energy, Metropolitan Energy Center
2020–2022
Co-PI (Missouri S&T PI)Electrifying Terminal Trucks in Unincentivized Markets
U.S. Department of Energy, Metropolitan Energy Center
2020–2022
PSU PIHow Bad is COVID-19 Hurting Electric Vehicle Usage and Mobility Patterns? A Case Study in Kansas City, Missouri for Agency and Industry Decision-Making
Google
2020
PIElectric Vehicle Charging Station (EVSE) Innovation: Streetlight Charging in City Right-of-Way
U.S. Department of Energy, Metropolitan Energy Center
2019–2022
Co-PI (Missouri S&T PI)Electric Vehicle Charging Station (EVSE) Innovation: Streetlight Charging in City Right-of-Way
U.S. Department of Energy, Metropolitan Energy Center
2019–2023
PSU PI
Recent publications
All 7 in this thrustPublic or Private? A Unified Demand–Supply Coupled Framework for EV Infrastructure Investment
VisionEV: Multimodal Large Language Models for Spatially Aware Electric Vehicle Charging Demand Prediction Using Satellite Imagery
Equitable Urban Electric Vehicle Charging: Feasibility and Benefits of Streetlight Charging in Kansas City Right-of-Way
Kansas City, Missouri, Streetlight Electric Vehicle Charging: Strategies and Challenges for Site Selection of Streetlight Electric Vehicle Infrastructure in Kansas City, Missouri
Learning-Based Demand-Supply-Coupled Charging Station Location Problem for Electric Vehicle Demand Management
Learning Electric Vehicle Driver Range Anxiety with an Initial State of Charge-Oriented Gradient Boosting Approach
Analysis of Electric Vehicle Charging Behavior Patterns with Function Principal Component Analysis Approach
Work with us
We help agencies and fleet operators turn EV and zero-emission truck data into decisions: where to put chargers, how to share investment, and how vehicles perform in real fleet service.
Email Dr. Hu