Comparison of Credit-Based Spending Models for Optimizing Express Lane Usage Under Incentives

Abstract: Express lanes are increasingly being considered for congestion management, with recent formulations of discounts and incentives aimed at equitable use of these lanes.This study introduces a new credit-based system where travelers earn credits by using public transit, which they can later spend to access express lanes. This shift from freely allocated credits to earned credits presents a novel challenge in understanding how such a scheme affects traveler behavior and system performance.

We use reinforcement learning to model the sequential credit-earning and spending behavior, incorporating key factors such as value of time, arrival delay penalties, departure times, and route choices. The model explicitly considers supply-side uncertainty and captures the stochastic variance in trip urgency across days, reflecting realistic travel conditions.

The study aims to investigate how travelers might adjust their behavior based on trip urgency, toll prices, and credit availability. For example, it will explore whether travelers with higher urgency deplete credits more quickly and whether lower toll prices increase the likelihood of express lane usage. Additionally, we will assess whether travelers, when low on credits, shift to transit use despite its lower perceived utility, and how this might incentivize transit adoption, ultimately benefiting overall traffic flow and equity.

This work seeks to contribute to the understanding of credit-based congestion pricing by offering insights into how earned credit schemes could optimize traveler satisfaction, encourage transit use when needed, and improve express lane efficiency.