Abstract This study introduces a comparative framework designed to evaluate the predictive capabilities of two distinct modeling approaches: traditional consideration and choice set generation models, and Machine Learning (ML) models. The framework focuses on a critical and often overlooked stage in decision–making process, which includes the set of alternatives that were considered prior to making a choice. To demonstrate the framework’s applicability, a synthetic route choice scenario was employed as a case study. In this context, routes with randomly assigned attributes were generated alongside synthetic populations of decision-makers, whose socioeconomic profiles were sampled from observed survey data. Each agent was assigned a set generation models to construct their individual consideration sets. ML models were then trained on these consideration sets to predict the considered alternatives of independent decision-maker populations. Two population scenarios were examined: one homogeneous, where all agents used the same set generation model, and one heterogeneous, where models were randomly assigned across agents. In both cases, the accuracy of the prediction was evaluated, along with a comparison between existing set generation and predictive models. The results showed that certain ML models can achieve over 80% accuracy in predicting individual considered items and over 60% accuracy in predicting entire consideration sets, across both population scenarios.

Before automated vehicles can safely operate in real-world traffic, it is crucial to ensure their reliability not only in normal conditions but also in rare and critical situations, such as traffic conflicts. Understanding these critical situations is essential for generating test cases that ensure robust system performance. However, current models of real-world traffic behavior in such situations are limited. This study addresses this gap by detecting rare critical situations at an urban signalized intersection, analyzing road user behavior, and deriving relevant parameter distributions through a long-term analysis of naturalistic trajectory data. Specifically, we focus on interactions between motorized road users (MRU) and crossing vulnerable road users (VRU) in illegal U-turn scenarios. Using over 180 days of video recordings, we extracted 9 million trajectories and identified four critical MRU–VRU interactions utilizing Surrogate Safety Measures and deceleration metrics. The analysis reveals that these interactions occur when the VRU traffic light switches from red to green. In addition, we descriptively model the driving behavior to generate parameter distributions for U-turn scenarios. Unlike previous studies, we differentiate between object classes, allowing us to effectively illustrate variations in curve radius—such as median values of 8.1 m for cars, 9.7 m for vans, and 14.3 m for trucks. Our results demonstrate an approach for modeling traffic participant behavior using large-scale trajectory data, showcasing a use case of data science in transportation and contributing valuable insights for simulation-based testing and scenario generation in automated vehicle development.

The substantial increase in traffic data offers new opportunities to inspect traffic congestion dynamics from different perspectives. This paper presents a novel framework for the interpretable representation and customizable retrieval of traffic congestion patterns using causal relation graphs, which harnesses many of these opportunities. By integrating domain knowledge with innovative data management techniques, we address the challenges of effectively handling and retrieving the growing volume of traffic data for diverse analytical purposes. The framework leverages causal graphs to encode traffic congestion patterns, capturing fundamental phenomena and their spatiotemporal relationships, thus facilitating an interpretable high-level view of traffic dynamics. Moreover, a customizable similarity measurement function is introduced based on inexact graph matching, allowing users to tailor the retrieval process to specific requirements. This framework’s capability to retrieve customizable and interpretable congestion patterns is demonstrated through extensive experiments with real-world highway traffic data in the Netherlands, highlighting its value in supporting diverse data-driven studies and applications.

MobilityCoins are a tradable mobility credit (TMC) scheme variant. TMC schemes are a cap-and-trade scheme for managing mobility that are designed to limit negative externalities, e.g., congestion, of traffic. Next to having link-specific or origin–destination-specific charges for cars as in the common TMC scheme, the MobilityCoin scheme’s distinctive elements are accommodating link-specific and origin-and-destination-specific charges and incentives for all modes of transport as well as being considered a mobility currency that can be earned, saved, and spent in multiple time periods. These distinctive features of the MobilityCoin scheme does not alter the core behavioral mechanism of TMC schemes of increasing car travel costs, but these features interfere with the credit market in terms of market volume and market price that ultimately affects traffic outcomes, e.g., an uncontrolled market volume increase can lower the market price that in turns increases the attractiveness of using the car. In this paper, we develop a mathematical model of multimodal macroscopic network flows and a MobilityCoin market to investigate the impacts of charges, incentives, and multi-period budgets. The model is implemented as a single-day model with an integration of sensitivity for multi-period budgets to study how the outcomes in the transportation system change with charges, incentives, and multi-period budgets. Further, we discuss implications for the policy design of MobilityCoins schemes.