PURPOSE:To assess the driving performance and visual attention of amblyopic drivers using a driving simulator integrated with eye tracking technology in an urban setting. METHODS:A total of 30 participants were enrolled: 14 amblyopic patients (all anisometropic and one also with strabismus) and 16 controls. Participants completed a simulated urban driving task while their eye movements and driving performance were recorded using a DriveSafety DS-600c simulator integrated with an infrared eye tracker. Driving performance (speed, brake reaction time / deceleration at pedestrian crossings, lane position variability, steering variability) and eye movement parameters (fixations, gaze distribution, saccades) were compared between amblyopic and control participants. In addition, pooling data from both groups, we assessed for correlations between eye movement parameters and driving performance. RESULTS:Amblyopic and nonamblyopic drivers performed similarly on driving tasks and had comparable eye movement recordings. Some nonsignificant tendencies were noted among amblyopic drivers, including slower driving speeds, longer brake reaction times, and increased steering variability. We identified potential tendencies for correlations between eye tracking metrics and driving performance: increased fixation frequency was associated with greater lane position variability, and larger saccadic movements correlated with improved lane stability and faster hazard response; however, these correlations were not statistically significant. CONCLUSIONS:In our study cohort, simulated driving performance and attention in subjects with anisometropic amblyopia under a standardized, relatively low-complexity urban scenario were similar to findings for nonamblyopic drivers, suggesting that amblyopia did not prevent safe driving in that context. Trends toward correlations between eye tracking and driving simulation parameters were observed, underscoring the potential of combining both approaches in future research in addition to testing a larger sample size, with a more challenging driving scenario.

Despite improvements in safety protocols, and evacuation planning, emergency situations due to various disasters still result in significant casualties. A key reason is the gap between recommended guidelines and the complexity of actual human behavior during stressful and panic-driven evacuations, which often leads to unsafe actions like stampedes. The study of human behavior in emergencies spans multiple fields such as mathematics, psychology, sociology, and engineering and aims to investigate how social interactions, environmental factors, and individual traits influence evacuation choices. Various mathematical models have been suggested to study this behavior. Nevertheless, many of these models lack integration of real-world data, which limits their accuracy and practical use. Recent technological advancements have improved the realism of research on evacuation dynamics, making it more reflective of actual disaster scenarios. These advancements stem primarily from the significant progress in data collection techniques and the rapid evolution of machine learning methods to utilise large datasets. This paper presents a comprehensive review of machine learning approaches trained on fixed datasets and applied to the study of evacuation dynamics. It traces the evolution of methodologies from conventional models to deep learning and generative frameworks, highlighting their capabilities and limitations in capturing real-world dynamics. The study systematically categorizes existing research based on model types, application domains, and data sources. Furthermore, this work investigates the relationship between model selection and dataset characteristics, emphasizing key trade-offs that influence model performance and applicability. It also identifies dominant research themes across different machine learning paradigms and evaluates how effectively these approaches address behavioral variability. Finally, the paper outlines the key challenges and proposes future research directions aimed at enhancing realism, scalability, and efficiency of the machine learning models in evacuation dynamics.

As urban planning policies and street design around the world are shifting towards more human-centric decisions and designs, it is crucial to explore and analyze schemes that aim at reducing the harm of car dependence in urban neighborhoods. This study develops a framework that defines the boundaries, prioritizes implementation, and assesses the traffic impact of low traffic neighborhoods (LTN). An LTN is a scheme where vehicular traffic is reduced in residential neighborhoods by mainly curbing through-traffic or “rat-running”. The boundaries are placed based on a set of criteria and assessed by locals and experts to fine tune them. Then, the LTNs are ranked in terms of implementation priority, relating to each LTN’s likelihood of succeeding and expected impact level based on a set of criteria. The implementability is assessed qualitatively by interviewing stakeholders, experts, and locals. The traffic impact is assessed using a traffic model. Before and after percentages of through-traffic on identified through-streets within LTNs are compared. This framework is then implemented in an area within Municipal Beirut with interventions of varying degrees of elimination of through-traffic in nine suggested LTNs based on set criteria of definition. The results show a decrease of through-traffic (up to complete elimination) in the neighborhoods where LTNs were implemented with minimal disruption to travel times to commuters still accessing destinations within the LTN.

Household travel surveys have been used for decades to collect individuals and households' travel behavior. However, self-reported surveys are subject to recall bias, as respondents might struggle to recall and report their activities accurately. This study examines the time reporting error of public transit users in a nationwide household travel survey by matching, at the individual level, five consecutive years of data from two sources, namely the Danish National Travel Survey (TU) and the Danish Smart Card system (Rejsekort). Survey respondents are matched with travel cards from the Rejsekort data solely based on the respondents' declared spatiotemporal travel behavior. Approximately, 70% of the respondents were successfully matched with Rejsekort travel cards. The findings reveal a median time reporting error of 11.34 minutes, with an Interquartile Range of 28.14 minutes. Furthermore, a statistical analysis was performed to explore the relationships between the survey respondents' reporting error and their socio-economic and demographic characteristics. The results indicate that females and respondents with a fixed schedule are in general more accurate than males and respondents with a flexible schedule in reporting their times of travel. Moreover, trips reported during weekdays or via the internet displayed higher accuracies compared to trips reported during weekends and holidays or via telephone interviews. This disaggregated analysis provides valuable insights that could help in improving the design and analysis of travel surveys, as well accounting for reporting errors/biases in travel survey-based applications. Furthermore, it offers valuable insights underlying the psychology of travel recall by survey respondents.

As municipalities are setting ambitious targets to increase non-motorized transportation mode shares, analytic frameworks for examining how such targets will be met become increasingly important. In this article, we update the Urban Network Analysis pedestrian modeling framework designed to link land use and urban development changes with pedestrian mobility outcomes by introducing new methods for capturing the effects of street properties on pedestrian travel demand and adjusting pedestrian trip generation rates with k-nearest accessibility scores based on destination availability in each context. This framework was used in conjunction with a participatory design process in Beirut, Lebanon to estimate pedestrian mobility impacts of three urban design scenarios. We illustrate how each scenario affects pedestrian trip generation across various trip types and trip distribution on individual street segments. Our approach demonstrates how urban design interventions–both land use changes and street quality improvements–can influence pedestrian travel demand. Estimates of these changes can provide planners and policymakers with valuable benefit-cost analyses of public space improvements, and a framework for understanding how site-specific planning and development decisions can impact progress towards (or away from) non-motorized mobility goals.

Sustainable communities are increasing in the Middle East and incorporate sustainable design elements like water reuse, urban farming, and green building design to economize resource consumption and keep greenhouse gas (GHG) emissions low. Sustainable transportation is perhaps the most challenging element to implement; it must be tailored to the location, size of the development, demographics, and prevalent culture. The literature on the sustainability and effectiveness of transportation aspects of sustainable communities in the Persian Gulf region is limited. To study these aspects in The Sustainable City (TSC), a 46-hectare mixed-use development housing 3,000 people near Dubai, we surveyed city residents and employees on their current travel behaviors and assessed how they respond to the availability of alternative mobility options. We found that, although residents want to live sustainably and actively attempt to do so, their transportation is sustainable only within the city's boundaries. For external trips, less sustainable options such as long-distance air travel, regular commuting to/from Dubai, and even short trips to neighboring communities drive up total transportation emissions. Residents are interested in shared mobility options, and an electric carsharing service could help reduce some travel impacts. However, the suburban setting of the development, fewer alternative mobility options currently available for trips to and from other locations in Dubai, a strong car culture, the lack of connectivity and integration challenges with neighboring communities, and the existing land-use patterns are major barriers to choosing sustainable transportation options.

Choice modelling is an increasingly important technique for forecasting and valuation, with applications in fields such as transportation, health and environmental economics. For this reason it has attracted attention from leading academics and practitioners and methods have advanced substantially in recent years. This Handbook, composed of contributions from senior figures in the field, summarises the essential analytical techniques and discusses the key current research issues. It will be of interest to academics, students and practitioners in a wide range of areas.

As cities are setting ambitious targets to increase non-motorized transportation mode share, analytic frameworks for examining how such targets will be met becomes increasingly important. In this article we update a pedestrian modeling framework designed to link land use changes, street renovations, and spatial development decisions to pedestrian mobility outcomes. This framework was used in conjunction with a participatory design process in Beirut, Lebanon to identify planning and urban design interventions that enhance walkability in neighborhoods most affected by the Beirut Port explosion of 2020. We estimate pedestrian mobility outcomes for three neighborhood development scenarios, illustrating how each one affects pedestrian trip generation across various trip types and trip distribution on individual street segments. Our approach demonstrates how urban design interventions–both land use changes and street quality improvements–can influence pedestrian travel demand. Estimates of these changes can provide planners and policymakers with valuable benefit-cost analyses of public space improvements.

In this article, a dynamic hybrid choice model of the driver stress arising from cognitive workload is developed and applied to a driving simulator experiment conducted with students at the American University of Beirut. A latent or unobserved variable quantifying the state stress over time is integrated with a discrete choice model of red-light violations. The driver state stress is induced by additional cognitive workload and situational factors in an urban driving context and is dependent on a time-invariant agent effect or individual trait. Driving performance (e.g., speed and acceleration) and physiological measures (e.g., heart rate) are used as indicators of the underlying state stress. The driver state stress is found to be significantly affected by road events or situations (e.g., encountering pedestrians, trucks, and traffic light), the level of cognitive workload, the individual propensity for stress, and the mere driving task (e.g., maintaining control of the car). Moreover, results show that there is a pattern of regulatory driving behavior in response to the increasing stress. The developed model is a mathematical conceptualization of the transactional model of the driver stress and can be integrated within in-vehicle systems to detect/predict the driver state and enhance safety and well-being.

We present a Gaussian Process - Latent Class Choice Model (GP-LCCM) to integrate a non-parametric class of probabilistic machine learning within discrete choice models (DCMs). Gaussian Processes (GPs) are kernel-based algorithms that incorporate expert knowledge by assuming priors over latent functions rather than priors over parameters, which makes them more flexible in addressing nonlinear problems. By integrating a Gaussian Process within a LCCM structure, we aim at improving discrete representations of unobserved heterogeneity. The proposed model would assign individuals probabilistically to behaviorally homogeneous clusters (latent classes) using GPs and simultaneously estimate class-specific choice models by relying on random utility models. Furthermore, we derive and implement an Expectation-Maximization (EM) algorithm to jointly estimate/infer the hyperparameters of the GP kernel function and the class-specific choice parameters by relying on a Laplace approximation and gradient-based numerical optimization methods, respectively. The model is tested on two different mode choice applications and compared against different LCCM benchmarks. Results show that GP-LCCM allows for a more complex and flexible representation of heterogeneity and improves both in-sample fit and out-of-sample predictive power. Moreover, behavioral and economic interpretability is maintained at the class-specific choice model level while local interpretation of the latent classes can still be achieved, although the non-parametric characteristic of GPs lessens the transparency of the model.

Human-made disasters continue to be the most dangerous of all types of disasters due to their suddenness and unpredictability. This calls for a thorough examination of people's evacuation behavior due to its impact on the evacuation procedure and its significance for evacuation planning. This study aims to provide a comprehensive understanding of the decision-making process of residents in case a human-made disaster occurs with an application to Beirut, Lebanon. This study is essential for pre-disaster planning, which mitigates potential damage from such disasters. Using structural equation modeling (SEM), the current study uses the Protective Action Decision Model (PADM) framework to explain intention toward evacuation behavior before a human-made disaster in three situations: being at home with all family members, having absent family members, and being at work or university when the event occurs. The findings of this study show that the PADM framework is relevant to explaining evacuation behavior intentions prior to a human-made disaster incident. Results indicate that knowledge perception does not trigger the intended behavior of evacuating immediately in the studied situations. Besides, the results indicate low confidence in the government's emergency plans and the unreliability of the official government warnings about human-made hazards. Overall, the findings of this study may contribute to a better understanding of evacuation behavior from disasters with less lead warning time. Besides, they may aid the Disaster Risk Management unit of Lebanon in developing emergency evacuation strategies that: understand the public's evacuation behavior; customize city-specific evacuation logistics; optimize the dissemination of evacuation information.

The aim of this study was to analyze the performance and attentional effects of sending voice messages while driving as compared to calling and texting. To this end, participants were asked to drive a given path while they either receive a phone call, send voice messages, or send text messages on a given cell phone, as well as a control condition. Driving performance, eye tracking, and subjective measures were collected. The results showed that voice messaging, while not as detrimental to driving performance as texting, does lead to similar levels of visual and cognitive distraction as texting and is generally more distracting than calling. Drivers also seem to be unaware of the dangers of voice messaging while driving. This research provides the basis for improved guidelines and legislation and more targeted awareness campaigns that emphasize the dangers of voice messaging while driving on a level with other banned practices.

This study presents a semi-nonparametric Latent Class Choice Model (LCCM) with a flexible class membership component. The proposed model formulates the latent classes using mixture models as an alternative approach to the traditional random utility specification with the aim of comparing the two approaches on various measures including prediction accuracy and representation of heterogeneity in the choice process. Mixture models are parametric model-based clustering techniques that have been widely used in areas such as machine learning, data mining and patter recognition for clustering and classification problems. An Expectation-Maximization (EM) algorithm is derived for the estimation of the proposed model. Using two different case studies on travel mode choice behavior, the proposed model is compared to traditional discrete choice models on the basis of parameter estimates' signs, value of time, statistical goodness-of-fit measures, and cross-validation tests. Results show that mixture models improve the overall performance of latent class choice models by providing better out-of-sample prediction accuracy in addition to better representations of heterogeneity without weakening the behavioral and economic interpretability of the choice models.

This case study investigates consumer preferences and stakeholders’ interests regarding hybrid electric vehicles (HEVs) and electric vehicles (EVs) in the Greater Beirut Area (GBA) in Lebanon where the market for these vehicle types is still nascent. A mixed logit model incorporating financial and technical attributes of common mid-size Internal Combustion Engines (ICE), HEVs, and EVs is estimated then used to compute Willingness to Pay (WTP) measures and evaluate the effectiveness of different monetary incentives in promoting these vehicle types. The study also uses qualitative research methods to capture the perspectives of different stakeholders regarding electric mobility in Lebanon. Data collected in 2018 through a stated preference survey reveals a WTP for a 100-km increase in driving range equal to 705 $ and for a 1 $ reduction in driving costs per 100 km equal to 305 $, values that are generally lower than several values found in the literature. A policy testing exercise suggests that doubling fuel taxes could increase the potential market shares of HEVs and EVs from 9.25% to 9.59% and from 4.98% to 5.84%, respectively. The provision of charging incentives to consumers could raise the market share of EVs up to 6.96%. A combination of both policies could further increase the proportion of EVs to 7.22%. In parallel, a stakeholder analysis draws attention to a multitude of challenges regarding the HEVs and EVs uptake as well as the public charging infrastructure rollout, namely excessive delays in establishing the enabling institutional and regulatory environment and shortcomings in the electricity supply. This research shows that it is more likely for HEVs than EVs to take off in the short term and that a solid transition to electric mobility in Lebanon necessitates further planning, especially in terms of instituting a clear and effective incentives scheme.

Car ownership and use is a main contributor to the deterioration of air quality in cities and to global warming. There is thus a pressing need to understand their determinants in this era of increasing demand for mobility. This paper studies car ownership and use decisions in a car-dominant developing country context, and quantifies the effect of public transportation availability on these decisions. A discrete–continuous modeling framework that estimates car ownership and use simultaneously is presented. People’s latent attitudes towards public transportation and the private car are also assumed to influence these decisions. The model is applied to the case of Lebanon, a developing country characterized by a high car ownership rate, a high percentage of trips made by car, and an informal public transportation system. Five policy scenarios involving potential improvements to the public transportation system, land use densification, or increase in fuel taxes were tested. The findings show that the current public transportation accessibility level has a minor impact on car ownership, but none on car usage. Only if major improvements to the public transportation services are enacted would a decrease in car ownership and usage be achieved. In such a case, model outcomes show that car ownership will be reduced in households with two cars by 5.88% and usage in general will be reduced by 15.22%. As a result, emissions and fuel consumption will be reduced by around 15%. Densification of zones outside Municipal Beirut is also a promising strategy for reducing car usage.

In this chapter, we present a methodological approach for Smart Mobility that integrates three key features: prediction, optimization, and personalization. They are integrated in such a way that when a travel menu is offered, predicted conditions are considered in the attributes of alternatives and optimized system-level policies are maintained. Similarly, user-level estimations and updates are used by prediction and optimization methods at the system-level in order to represent the population with most up-to-date behavioral estimates. Furthermore, a simulation-based evaluation methodology enables to validate the performance of prediction, optimization, and personalization before Smart Mobility is implemented in real-life. Two case studies are presented based on the proposed methodologies together with platforms that facilitate their application. Potential benefits of the proposed methodologies are evaluated which can be classified into user-level and system-level benefits. User-level benefits include consumer surplus, waiting times, etc., and system-level is concerned with congestion, throughput, system-wide travel time, etc. As there is normally a tradeoff between the individual decision-making and system-wide decision-making, Smart Mobility bridges them together with appropriate methodologies on each end. For example, for our Flexible Mobility on Demand case study, we observe 10%–20% reduction in volume-to-capacity ratio as a system-level benefit. Moreover, we see that the tradeoff between consumer surplus and operator profit can be managed with an appropriate objective function.

This study uses a natural experiment in Beirut, Lebanon, to investigate the effects of a street-level urban design intervention that improved the walking environment through a wider sidewalk, removal of a parking lane, raised junctions, and other elements. This study analyzes the impacts on pedestrian flow, pedestrian satisfaction with the walking experience, commercial activity, and business managers’ attitudes. Difference-in-difference regressions suggest that the main effect of such interventions is not necessarily an increase in pedestrian traffic, but instead safer pedestrian maneuvering and a better walking experience. It is also found through descriptive analysis that while businesses and shops experience increased business post-intervention, noticeable dissatisfaction with the intervention is reported by managers and owners. It is hypothesized that this dissatisfaction is a result of the lengthy construction process renovating and refurbishing the street, and the removal of parking spaces. Policy recommendations are drawn for the mitigation of business managers’ concerns and the enhancement of the walking environment for the design of future similar interventions.

Ridesourcing (Uber, Careem, Lyft, …) is emerging as a main player in the transportation industry. However, its relation to mass transit remains ambiguous, with divided opinions on its complementarity or substitutive effect towards high capacity public transportation systems. This study examines the integration of ridesourcing and transit, particularly focusing on modeling the demand for mass transit when ridesourcing is used as an access or egress mode to mass transit. It extends the existing literature on the integration of transit and new mobility concepts by providing a modeling framework that incorporates all stages of multi-modal trips such as those that involve using mass transit. A mixed logit with error component structure is presented to capture correlations in unobserved factors across multi-modal alternatives sharing similar modes at certain stages. The framework incorporates uni-modal and multi-modal travel alternatives and distinguishes between access, main mode, and egress stages without applying constraints on possible combinations. An application to Beirut’s planned Bus Rapid Transit (BRT) system, performed on a data set of 392 respondents, reveals that ridesourcing as a feeder mode is mostly popular with young commuters while also being perceived as more reliable than feeder buses and jitneys. Awareness and familiarity are major drivers for the service implying higher potential in the future. A complementarity effect with transit is found as the introduction of ridesourcing at the feeders’ level is expected to drive an additional 2% of commuters to use the BRT. Decreasing ridesourcing fare is effective for its integration with transit, as a fare decrease of 50% increases BRT market share from 33.53% to 36.89% of all motorized trips, implying possible synergies between the two modes. Forecasting results further reveal that additional taxes on parking used by car commuters and increasing park and ride capacity at BRT stations are effective policies to augment BRT ridership.

We develop a methodology to analyze pedestrian-vehicular interactions in urban streets in a mixed traffic environment, and then apply it to Bliss Street, an urban street in Beirut. Data on the street was collected before and after a crosswalk was installed using videography, radar speed guns, and manual counts. A pedestrian gap acceptance model indicated that installing the crosswalk did not have any significant effect on the pedestrians’ sensitivity to waiting time, gap size, or the speed of the approaching vehicles. However, it caused reductions in the speed of approaching vehicles which in turn encouraged pedestrians to accept shorter gaps. A micro-simulation model indicated that the crosswalk would reduce the speed on the street slightly, with significant reductions observed if more pedestrians who currently cross at midblock locations shift to use the crosswalk. The results of this study can be used to test interventions for enhancing pedestrian safety in Lebanon, and are generalizable to similar contexts in developing countries.

This study investigates the potential market demand of shared-ride taxi and shuttle services designed to serve members of organizations in dense urbanized areas. It develops and compares two different multivariate count data modeling approaches, the multinomial distribution and the full enumeration of count alternatives, under an integrated choice and latent variable framework. The study accounts for day-to-day variability in commuting behavior, also known as multimodality, by modeling the weekly frequency of commuting by different travel modes instead of modeling choices for a single trip/day. Using stated preference data collected in the Spring of Academic Year 2016–2017, the models are applied to a case study of students who are highly dependent on private cars at the American University of Beirut (AUB), Lebanon. Policy analysis is conducted to investigate the impact of different price levels and modal attributes on the students’ mode choice behavior. Under practical scenarios, results show that more than 55% of students would adopt a multimodal travel behavior in a given week and that 9–20% of trips are expected to be made by shared-taxi and 12–25% by shuttle. Thus, modeling single trip/day choices instead of weekly decisions would lead to limitations in model forecasts related to the full impact of the proposed policies over longer periods. Results also show that the full enumeration model guarantees higher prediction accuracy and results in an estimate of value of time that is closer to other local estimates for the study area.