Journal Article
Demand Modeling for Advanced Air Mobility: Challenges, Opportunities, and Future Directions
A comprehensive IEEE T-ITS survey on Advanced Air Mobility demand modeling, covering econometric, discrete choice, simulation, machine learning, and hybrid approaches for future air transportation systems.
Abstract
Advanced Air Mobility (AAM) has attracted increasing research and industry attention as a potential extension of passenger and cargo transportation systems, supported by developments in electric vertical take-off and landing (eVTOL) aircraft, autonomy, and integrated traffic management. Unlike conventional transportation systems, AAM presents unique challenges to demand modeling due to untested routes, new vehicle classes, and evolving user behavior. This paper conducts a comprehensive survey of demand modeling approaches relevant to AAM, encompassing econometric models, discrete choice frameworks, simulation-based approaches, machine learning techniques, and hybrid methods. Our evaluation focuses on four core modeling stages: trip generation, trip distribution, mode choice, and assignment. We further analyze the key factors influencing demand and assess their implications for model design and forecasting accuracy. Research gaps are identified in areas such as airspace routing, vertiport placement, capacity constraints, weather resilience, and community acceptance. To address these, we highlight the need for infrastructure-aware, behaviorally grounded, and real-time adaptive demand models. We outline future research opportunities including fine-grained temporal and spatial resolution, multimodal integration, induced demand representation, and equity considerations. By categorizing current methods and synthesizing emerging directions, this survey establishes a critical research agenda to guide policymakers, practitioners, and researchers in advancing robust and socially responsible AAM demand forecasting.
In brief
What This Work Does
This paper explains how researchers and planners can estimate future demand for air taxi, regional air mobility, and related AAM services when the market, routes, infrastructure, and user behavior are still emerging.
Research impact
Why It Matters
Advanced Air Mobility cannot be planned only around aircraft technology. Its success depends on whether travelers will use the service, where demand will emerge, how it will interact with existing transportation modes, and whether infrastructure such as vertiports, charging systems, and airspace corridors can support that demand. This paper provides a structured foundation for researchers, planners, and policymakers to build more realistic AAM demand forecasting models.
Survey at a glance
A Structured View of AAM Demand
Four-Step AAM Demand Modeling Framework
Trip Generation
Identifies potential AAM trips from existing travel patterns, commuting flows, airport-access trips, regional travel demand, cargo movement, and emergency service needs.
Trip Distribution
Estimates spatial origin-destination flows and candidate AAM corridors using gravity models, OD matrices, spatial interaction models, and scenario-based assumptions.
Mode Choice
Estimates how travelers may choose AAM over cars, public transit, ride-hailing, or conventional air travel based on time, cost, access, reliability, safety perception, and willingness to pay.
Assignment
Allocates AAM trips to routes, vertiports, time periods, and operational networks while considering fleet availability, capacity, weather, charging, and airspace constraints.
Key Contributions
- Surveys major methodological families used for AAM demand modeling.
- Maps AAM demand research to trip generation, distribution, mode choice, and assignment.
- Identifies infrastructure, behavioral, operational, and equity gaps in current forecasting practice.
- Develops a research agenda for robust and socially responsible AAM demand forecasting.
Modeling Approaches Reviewed
Econometric Models
Useful for strategic demand estimation using socioeconomic, geographic, and travel variables.
Discrete Choice Models
Useful for modeling traveler decisions and modal split between AAM and competing modes.
Simulation-Based Models
Useful for testing operational feasibility, vertiport congestion, fleet scheduling, and network performance.
Machine Learning Models
Useful for data-driven forecasting, demand hotspot detection, clustering, and adaptive prediction.
Hybrid Models
Promising for integrating behavior, infrastructure, vehicle performance, and operational constraints into one modeling framework.
Research Gaps
Infrastructure
Vertiport placement · Capacity constraints
Operations
Airspace routing · Weather resilience
Traveler Behavior
Community acceptance · Induced demand
System Design
Multimodal integration · Equity
Who this survey is for
Designed for Research and Decision-Making
Researchers
Developing demand-forecasting, behavioral, simulation, machine-learning, and hybrid methods.
Transportation Planners
Evaluating vertiports, networks, capacity, multimodal connections, and operational scenarios.
Policymakers
Considering access, equity, community acceptance, resilience, and socially responsible deployment.
Publication Details
- Type
- Journal Article
- Venue
- IEEE Transactions on Intelligent Transportation Systems
- Year
- 2026
- Published
- March 16, 2026
- Pages
- 1-21
Research Topics
Links and Access
Citation
@ARTICLE{11435518,
author={Acharya, Kamal and Raza, Waleed and Vasiloff, Katherine and Wang, Zhenbo and Sun, Liang and Song, Houbing Herbert},
journal={IEEE Transactions on Intelligent Transportation Systems},
title={Demand Modeling for Advanced Air Mobility: Challenges, Opportunities, and Future Directions},
year={2026},
volume={},
number={},
pages={1-21},
keywords={Active appearance model;Atmospheric modeling;Urban air mobility;Random access memory;Aircraft;Surveys;Biological system modeling;Aircraft propulsion;Adaptation models;Analytical models;Advanced air mobility;air transportation;data analytics and data science;transportation demand modeling and estimation},
doi={10.1109/TITS.2026.3671002}}