Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Integrating Travel Behavior Forecasting and Generative Modeling for Predicting Future Urban Mobility and Spatial Transformations

Eugene Denteh, Andrews Danyo, Joshua Kofi Asamoah, Blessing Agyei Kyem, Twitchell Addai, Armstrong Aboah

TL;DR

The paper tackles the challenge of forecasting long-term urban mobility and its spatial manifestations for proactive planning. It introduces a two-stage framework that first uses a Temporal Fusion Transformer to forecast travel-behavior from demographic data, then conditions a Generative Adversarial Network to synthesize future satellite imagery reflecting those mobility trends. Key contributions include injecting low-dimensional tabular data into a GAN for high-fidelity spatial predictions, and demonstrating improved temporal accuracy (e.g., RMSE, $R^2$, DTW) along with realistic spatial synthesis (FID, SSIM). The work offers a data-driven, integrative tool for transportation planning that can reduce infrastructure wastage, support government systems, and enable scenario-based urban development.

Abstract

Transportation planning plays a critical role in shaping urban development, economic mobility, and infrastructure sustainability. However, traditional planning methods often struggle to accurately predict long-term urban growth and transportation demands. This may sometimes result in infrastructure demolition to make room for current transportation planning demands. This study integrates a Temporal Fusion Transformer to predict travel patterns from demographic data with a Generative Adversarial Network to predict future urban settings through satellite imagery. The framework achieved a 0.76 R-square score in travel behavior prediction and generated high-fidelity satellite images with a Structural Similarity Index of 0.81. The results demonstrate that integrating predictive analytics and spatial visualization can significantly improve the decision-making process, fostering more sustainable and efficient urban development. This research highlights the importance of data-driven methodologies in modern transportation planning and presents a step toward optimizing infrastructure placement, capacity, and long-term viability.

Integrating Travel Behavior Forecasting and Generative Modeling for Predicting Future Urban Mobility and Spatial Transformations

TL;DR

The paper tackles the challenge of forecasting long-term urban mobility and its spatial manifestations for proactive planning. It introduces a two-stage framework that first uses a Temporal Fusion Transformer to forecast travel-behavior from demographic data, then conditions a Generative Adversarial Network to synthesize future satellite imagery reflecting those mobility trends. Key contributions include injecting low-dimensional tabular data into a GAN for high-fidelity spatial predictions, and demonstrating improved temporal accuracy (e.g., RMSE, , DTW) along with realistic spatial synthesis (FID, SSIM). The work offers a data-driven, integrative tool for transportation planning that can reduce infrastructure wastage, support government systems, and enable scenario-based urban development.

Abstract

Transportation planning plays a critical role in shaping urban development, economic mobility, and infrastructure sustainability. However, traditional planning methods often struggle to accurately predict long-term urban growth and transportation demands. This may sometimes result in infrastructure demolition to make room for current transportation planning demands. This study integrates a Temporal Fusion Transformer to predict travel patterns from demographic data with a Generative Adversarial Network to predict future urban settings through satellite imagery. The framework achieved a 0.76 R-square score in travel behavior prediction and generated high-fidelity satellite images with a Structural Similarity Index of 0.81. The results demonstrate that integrating predictive analytics and spatial visualization can significantly improve the decision-making process, fostering more sustainable and efficient urban development. This research highlights the importance of data-driven methodologies in modern transportation planning and presents a step toward optimizing infrastructure placement, capacity, and long-term viability.

Paper Structure

This paper contains 41 sections, 32 equations, 6 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Historical Challenges in Transportation Infrastructure Planning and Their Socio-Spatial Consequences
  4. Challenges in Predicting High-Dimensional Spatial Outcomes from Low-Dimensional Data
  5. Data
  6. Methodology
  7. Data Preprocessing
  8. Temporal Fusion Transformer for Travel Behavior Prediction
  9. Problem Formulation
  10. Encoder-Decoder Architecture
  11. Multi-Head Attention Mechanism
  12. Feed-Forward Network
  13. Loss Function and Optimization
  14. Generative Adversarial Network for Spatial Prediction
  15. Tabular Data Injection
  16. ...and 26 more sections

Figures (6)

  • Figure 1: Overview of the proposed framework for the Temporal Fusion Transformer
  • Figure 2: Overview of the proposed framework for the Generative Adversarial Network
  • Figure 3: Actual Target (Travel Behavior) Feature Trends Over Years
  • Figure 4: Attention Weights Visualization. The attention mechanism assigns varying importance to different time steps, but fails to emphasize the critical period leading up to 2020.
  • Figure 5: Predicted Target Feature Trends (2018 - 2023) using the TFT Model. The model successfully captures the overall trend in travel behavior, demonstrating its effectiveness.
  • ...and 1 more figures