A practical method to estimate the benefits of improved road network reliability: an application to departing air passengers

This paper develops and applies a practical method to estimate the benefits of improved reliability of road networks. We present a general methodology to estimate the scheduling costs due to travel time variability for car travel. In contrast to existing practical methods, we explicitly consider the effect of travel time variability on departure time choices. We focus on situations when only mean delays are known, which is typically the case when standard transport models are used. We first show how travel time variability can be predicted from mean delays. We then estimate the scheduling costs of travellers, taking into account their optimal departure time choice given the estimated travel time variability. We illustrate the methodology for air passengers traveling by car to Amsterdam Schiphol Airport. We find that on average planned improvements in network reliability only lead to a small reduction in access costs per trip in absolute terms, mainly because most air passengers drive to the airport outside peak hours, when travel time variability tends to be low. However, in relative terms the reduction in access costs due to the improvements in network reliability is substantial. In our case we find that for every 1 Euro reduction in travel time costs, there is an additional cost reduction of 0.7 Euro due to lower travel time variability, and hence lower scheduling costs. Ignoring the benefits from improved reliability may therefore lead to a severe underestimation of the total benefits of infrastructure improvements.     

Latent temporal preferences: An application to airline travel

An essential element of demand modeling in the airline industry is the representation of time of day demand—the demand for a given itinerary as a function of its departure or arrival times. It is an important datum that drives successful scheduling and fleet decisions. There are two key components to this problem: the distribution of the time of day demand and how preferred travel time influences itinerary choice. This paper focuses on estimating the time of day distribution. Our objective is to estimate it in a manner that is not confounded with air travel supply; is a function of the characteristics of the traveler, the trip, and the market; and accounts for potential measurement errors in self-reported travel time preferences. We employ a stated preference dataset collected by intercepting people who were booking continental US trips via an internet booking service. Respondents reported preferred travel times as well as choices from a hypothetical set of itineraries. We parameterize the time of day distribution as a mixture of normal distributions (due to the strong peaking nature of travel time preferences) and allow the mixing function to vary by individual characteristics and trip attributes. We estimate the time of day distribution and the itinerary choice model jointly in a manner that accounts for measurement error in the self-reported travel time preferences. We find that the mixture of normal distributions fits the time of day distribution well and is behaviorally intuitive. The strongest covariates of travel time preferences are party size and time zone change. The methodology employed to treat self-reported travel time preferences as potentially having error contributes to the broader transportation time of day demand literature, which either assumes that the desired travel times are known with certainty or that they are unknown. We find that the error in self-reported travel time preferences is statistically significant and impacts the inferred time of day demand distribution.     

Travel time ratio: the key factor of spatial reach

An important aspect of reach and accessibility is the time people are willing to spend on reaching activity places. In this paper we see the issue of travel time in an alternative way. Instead of looking at travel time separated from time spent on activities, we examine the relation between travel time and stay time. We operationalize this relation with the concept “travel time ratio”. A hypothetical framework underlying these travel time ratios is displayed. We show that for similar types of activity places the value of travel time ratio are in accordance with each other. We find large differences between trips for mandatory activities and trips for discretionary activities. The results indicate the stability of the travel time ratios. Finally, some implications for future research and policy will be mentioned. This revised version was published online in June 2006 with corrections to the Cover Date.     

A bi-objective user equilibrium model of travel time reliability in a road network

Travel time, travel time reliability and monetary cost have been empirically identified as the most important criteria influencing route choice behaviour. We concentrate on travel time and travel time reliability and review two prominent user equilibrium models incorporating these two factors. We discuss some shortcomings of these models and propose alternative bi-objective user equilibrium models that overcome the shortcomings. Finally, based on the observation that both models use standard deviation of travel time within their measure of travel time reliability, we propose a general travel time reliability bi-objective user equilibrium model. We prove that this model encompasses those discussed previously and hence forms a general framework for the study of reliability related user equilibrium. We demonstrate and validate our concepts on a small three-link example.     

Door-to-door travel times in RP departure time choice models: An approximation method using GPS data

A common way to determine values of travel time and schedule delay is to estimate departure time choice models, using stated preference (SP) or revealed preference (RP) data. The latter are used less frequently, mainly because of the difficulties to collect the data required for the model estimation. One main requirement is knowledge of the (expected) travel times for both chosen and unchosen departure time alternatives. As the availability of such data is limited, most RP-based scheduling models only take into account travel times on trip segments rather than door-to-door travel times, or use very rough measures of door-to-door travel times. We show that ignoring the temporal and spatial variation of travel times, and, in particular, the correlation of travel times across links may lead to biased estimates of the value of time (VOT). To approximate door-to-door travel times for which no complete measurement is possible, we develop a method that relates travel times on links with continuous speed measurements to travel times on links where relatively infrequent GPS-based speed measurements are available. We use geographically weighted regression to estimate the location-specific relation between the speeds on these two types of links, which is then used for travel time prediction at different locations, days, and times of the day. This method is not only useful for the approximation of door-to-door travel times in departure time choice models, but is generally relevant for predicting travel times in situations where continuous speed measurements can be enriched with GPS data.     

Discretised link travel time models based on cumulative flows: Formulations and properties

In the research area of dynamic traffic assignment, link travel times can be derived from link cumulative inflow and outflow curves which are generated by dynamic network loading. In this paper, the profiles of cumulative flows are piecewise linearized. Both the step function (SF) and linear interpolation (LI) are used to approximate cumulative flows over time. New formulations of the SF-type and LI-type link travel time models are developed. We prove that these two types of link travel time models ensure first-in-first-out (FIFO) and continuity of travel times with respect to flows, and have other desirable properties. Since the LI-type link travel time model does not satisfy the causality property, a modified LI-type (MLI-type) link travel time model is proposed in this paper. We prove that the MLI-type link travel time model ensures causality, strong FIFO and travel time continuity, and that the MLI-type link travel time function is strictly monotone under the condition that the travel time of each vehicle on a link is greater than the free flow travel time on that link. Numerical examples are set up to illustrate the properties and accuracy of the three models.     

Accounting for travel time variability in the optimal pricing of cars and buses

A number of studies have shown that in addition to travel time and cost as the common influences on mode, route and departure time choices, travel time variability plays an increasingly important role, especially in the presence of traffic congestion on roads and crowding on public transport. The dominant focus of modelling and implementation of optimal pricing that incorporates trip time variability has been in the context of road pricing for cars. The main objective of this paper is to introduce a non-trivial extension to the existing literature on optimal pricing in a multimodal setting, building in the role of travel time variability as a source of disutility for car and bus users. We estimate the effect of variability in travel time and bus headway on optimal prices (i.e., tolls for cars and fares for buses) and optimal bus capacity (i.e., frequencies and size) accounting for crowding on buses, under a social welfare maximisation framework. Travel time variability is included by adopting the well-known mean–variance model, using an empirical relationship between the mean and standard deviation of travel times. We illustrate our model with an application to a highly congested corridor with cars, buses and walking as travel alternatives in Sydney, Australia. There are three main findings that have immediate policy implications: (i) including travel time variability results in higher optimal car tolls and substantial increases in toll revenue, while optimal bus fares remain almost unchanged; (ii) when bus headways are variable, the inclusion of travel time variability as a source of disutility for users yields higher optimal bus frequencies; and (iii) including both travel time variability and crowding discomfort leads to higher optimal bus sizes.     

Predicting travel time reliability using mobile phone GPS data

Estimates of road speeds have become commonplace and central to route planning, but few systems in production provide information about the reliability of the prediction. Probabilistic forecasts of travel time capture reliability and can be used for risk-averse routing, for reporting travel time reliability to a user, or as a component of fleet vehicle decision-support systems. Many of these uses (such as those for mapping services like Bing or Google Maps) require predictions for routes in the road network, at arbitrary times; the highest-volume source of data for this purpose is GPS data from mobile phones. We introduce a method (TRIP) to predict the probability distribution of travel time on an arbitrary route in a road network at an arbitrary time, using GPS data from mobile phones or other probe vehicles. TRIP captures weekly cycles in congestion levels, gives informed predictions for parts of the road network with little data, and is computationally efficient, even for very large road networks and datasets. We apply TRIP to predict travel time on the road network of the Seattle metropolitan region, based on large volumes of GPS data from Windows phones. TRIP provides improved interval predictions (forecast ranges for travel time) relative to Microsoft’s engine for travel time prediction as used in Bing Maps. It also provides deterministic predictions that are as accurate as Bing Maps predictions, despite using fewer explanatory variables, and differing from the observed travel times by only 10.1% on average over 35,190 test trips. To our knowledge TRIP is the first method to provide accurate predictions of travel time reliability for complete, large-scale road networks.     

Eco-routing problem considering fuel consumption and probabilistic travel time budget

Road transportation is one of the major sources of greenhouse gas emissions. To reduce energy consumption and alleviate this environmental problem, this study aims to develop an eco-routing algorithm for navigation systems. Considering that both fuel consumption and travel time are important factors when planning a trip, the proposed routing algorithm finds a path that consumes the minimum amount of gasoline while ensuring that the travel time satisfies a specified travel time budget and an on-time arrival probability. We first develop link-based fuel consumption models based on vehicle dynamics, and then the Lagrangian-relaxation-based heuristic approach is proposed to efficiently solve this NP-hard problem. The performance of the proposed eco-routing strategy is verified in a large-scale network with real travel time and fuel consumption data. Specifically, a sensitivity analysis of fuel consumption reduction for travel demand and travel time buffer is discussed in our simulation study.     

Link travel time inference using entry/exit information of trips on a network

This paper studies link travel time estimation using entry/exit time stamps of trips on a steady-state transportation network. We propose two inference methods based on the likelihood principle, assuming each link associates with a random travel time. The first method considers independent and Gaussian distributed link travel times, using the additive property that trip time has a closed-form distribution as the summation of link travel times. We particularly analyze the mean estimates when the variances of trip time estimates are known with a high degree of precision and examine the uniqueness of solutions. Two cases are discussed in detail: one with known paths of all trips and the other with unknown paths of some trips. We apply the Gaussian mixture model and the Expectation–Maximization (EM) algorithm to deal with the latter. The second method splits trip time proportionally among links traversed to deal with more general link travel time distributions such as log-normal. This approach builds upon an expected log-likelihood function which naturally leads to an iterative procedure analogous to the EM algorithm for solutions. Simulation tests on a simple nine-link network and on the Sioux Falls network respectively indicate that the two methods both perform well. The second method (i.e., trip splitting approximation) generally runs faster but with larger errors of estimated standard deviations of link travel times.     

Travel time variability and airport accessibility

We analyze the cost of access travel time variability for air travelers. Reliable access to airports is important since the cost of missing a flight is likely to be high. First, the determinants of the preferred arrival times at airports are analyzed. Second, the willingness to pay (WTP) for reductions in access travel time, early and late arrival time at the airport, and the probability to miss a flight are estimated, using a stated choice experiment. The results indicate that the WTPs are relatively high. Third, a model is developed to calculate the cost of variable travel times for representative air travelers going by car, taking into account travel time cost, scheduling cost and the cost of missing a flight using empirical travel time data. In this model, the value of reliability for air travelers is derived taking “anticipating departure time choice” into account, meaning that travelers determine their departure time from home optimally. Results of the numerical exercise show that the cost of access travel time variability for business travelers are between 0% and 30% of total access travel cost, and for non-business travelers between 0% and 25%. These numbers depend strongly on the time of the day.     

On the cost of misperceived travel time variability

Because individuals may misperceive travel time distributions, using the implied reduced form of the scheduling model might fall short of capturing all costs of travel time variability. We reformulate a general scheduling model employing rank-dependent utility theory and derive two special cases as econometric specifications to study these uncaptured costs. It is found that reduced-form expected cost functions still have a mean–variance form when misperception is considered, but the value of travel time variability is higher. We estimate these two models with stated-preference data and calculate the empirical cost of misperception. We find that: (i) travelers are mostly pessimistic and thus tend to choose departure times too early to achieve a minimum cost, (ii) scheduling preferences elicited using a stated-choice method can be relatively biased if probability weighting is not considered, and (iii) the extra cost of misperceiving the travel time distribution might be nontrivial when time is valued differently over the time of day and is substantial for some people.     

Equilibrium analysis of the scheduling of tours in congested networks

In this paper, we develop a model of travel in tours that joins several locations by travel through a congested network. We develop a microscopic analysis in continuous time of individual benefits obtained by spending time at each of the locations and costs incurred through travel between them. This is combined with a continuous time macroscopic equilibrium model of travel during congested peak periods to show how individuals' travel choices are influenced by the congestion that result from corresponding choices made by others. We show how different travellers can achieve identical net utilities by making different combinations of choices within the equilibrium. The resulting model can be used to investigate the effect on travel behaviour and individual utility of various transport interventions, and we illustrate this by considering the effect of a peak‐period charge that eliminates congestion.     

Route choice decision under travel time uncertainty

We study route choice behavior when travel time is uncertain. In this case, users choice depends both on expected travel time and travel time variability. We collected survey data in the Paris area and analyzed them using a method based on the ordered probit. This leads to an ordinal as well as to different cardinal measures of risk aversion. Such an approach is consistent with expected and with non-expected utility theory. Econometric estimates suggest that absolute risk aversion is constant and show that risk aversion is larger for transit users, blue collars and for business appointments.     

Traveler delay costs and value of time with trip chains, flexible activity scheduling and information

The delay costs of traffic disruptions and congestion and the value of travel time reliability are typically evaluated using single trip scheduling models, which treat the trip in isolation of previous and subsequent trips and activities. In practice, however, when activity scheduling to some extent is flexible, the impact of delay on one trip will depend on the actual and predicted travel time on itself as well as other trips, which is important to consider for long-lasting disturbances and when assessing the value of travel information. In this paper we extend the single trip approach into a two trips chain and activity scheduling model. Preferences are represented as marginal activity utility functions that take scheduling flexibility into account. We analytically derive trip timing optimality conditions, the value of travel time and schedule adjustments in response to travel time increases. We show how the single trip models are special cases of the present model and can be generalized to a setting with trip chains and flexible scheduling. We investigate numerically how the delay cost depends on the delay duration and its distribution on different trips during the day, the accuracy of delay prediction and travel information, and the scheduling flexibility of work hours. The extension of the model framework to more complex schedules is discussed.     

Influence of peak spreading and departure time choice on evaluation of its measures

Intelligent transport systems provide various means to improve capacity and travel time in road networks. Evaluation of the benefits of these improvements requires consideration of travellers' response to them. We consider a continuous‐time equilibrium model of departure time choice and identify a formula for the dynamic equilibrium departure rate profile. We develop the analysis to consider the effect on the cost incurred by travellers of ITS measures through their effects on each of the travel time in the absence of congestion, and the capacity for travel. This shows the importance in choice of departure time of travellers' values of time at each of the origin and destination of their journeys. We show the importance of these values of time in evaluation, and that if travellers value their time at both the origin and destination of their journeys, their responses will lead them to achieve a greater reduction in costs than would be achieved under free‐flow conditions.     

Proactive route guidance to avoid congestion

We propose a proactive route guidance approach that integrates a system perspective: minimizing congestion, and a user perspective: minimizing travel inconvenience. The approach assigns paths to users so as to minimize congestion while not increasing their travel inconvenience too much. A maximum level of travel inconvenience is ensured and a certain level of fairness is maintained by limiting the set of considered paths for each Origin-Destination pair to those whose relative difference with respect to the shortest (least-duration) path, called travel inconvenience, is below a given threshold. The approach hierarchically minimizes the maximum arc utilization and the weighted average experienced travel inconvenience. Minimizing the maximum arc utilization in the network, i.e., the ratio of the number of vehicles entering an arc per time unit and the maximum number of vehicles per time unit at which vehicles can enter the arc and experience no slowdown due to congestion effects, is a system-oriented objective, while minimizing the weighted average experienced travel inconvenience, i.e., the average travel inconvenience over all eligible paths weighted by the number of vehicles per time unit that traverse the path, is a user-oriented objective. By design, to ensure computational efficiency, the approach only solves linear programming models. In a computational study using benchmark instances reflecting a road infrastructure encountered in many cities, we analyze, for different levels of maximum travel inconvenience and, the minimum maximum arc utilization and the weighted average experienced travel inconvenience. We find that accepting relatively small levels of maximum travel inconvenience can result in a significant reduction, or avoiding, of congestion.     

Tradeoffs between Time Allocations to Maintenance Activities/Travel and Discretionary Activities/Travel

This paper focuses on the tradeoff in time allocation between maintenance activities/travel and discretionary activities/travel. We recognize that people generally must travel a minimum amount of time in order to allocate one unit of time to the activity. This minimum amount of travel is represented by the travel time price, a ratio obtained by dividing the total amount of time traveling to maintenance or discretionary activities by the total amount of time spent on activities of the same type; it is the time equivalent of the monetary price for performing an activity. Using the San Francisco Bay Area 1996 Household Travel Survey data and applying the Almost Ideal Demand System (AIDS) of demand equations, we found that with respect to the time equivalent of income elasticities of maintenance and discretionary activities, the former is less than unity and the latter is greater than unity. In other words, maintenance activities are a necessity and discretionary activities are a luxury. With respect to the own travel time price elasticities, if the travel time price of performing a certain type of activity increases (for reasons such as traffic congestion), one would reduce the time allocated to that type of activity. Time spent on maintenance activities is less elastic than the time spent on discretionary activities. As for the cross travel time price elasticities (changes in time allocated to activity type i in responses to changes in the time price for activity type j), we found that ɛ_dm>0 and ɛ_md>0, suggesting a substitution effect between maintenance and discretionary activities.     

Not All Transfers Are Created Equal: Towards a Framework Relating Transfer Connectivity to Travel Behaviour

Abstract

Walking from origins to transit stops, transferring between transit lines and walking from transit stops to destinations—all add to the burden of transit travel, sometimes to a very large degree. Transfers in particular can be stressful and/or time‐consuming for travellers, discouraging transit use. As such, transit facilities that reduce the burdens of walking, waiting and transferring can substantially increase transit system efficacy and use. In this paper, we argue that transit planning research on transit stops and stations, and transit planning practice frequently lack a clear conceptual framework relating transit waits and transfers with what we know about travel behaviour. Therefore, we draw on the concepts of transfer penalties and value of time in the travel behaviour/economics literature to develop a framework that situates transfer penalties within the total travel generalized costs of a transit trip. For example, value of time is important in relating actual time of waiting and walking to the perceived time of travel. We also draw on research to classify factors most important to users’ perspectives and travel behaviour—transfer costs, time scheduling and five transfer facility attributes: (1) access, (2) connection and reliability, (3) information, (4) amenities, and (5) security and safety. Using this framework, we seek to explicitly relate improvements of transfer stops/stations with components of transfer penalties and changes in travel behaviour (through a reduction in transfer penalties). We conclude that the employment of such a framework can help practitioners better apply the most effective improvements to transit stops and transfer facilities.     

Value of time – A citizen perspective

The dominant empirical approach to infer Value of Time is based on experiments in which respondents are typically asked to make hypothetical travel choices as if they were paying travel costs from their own budget, in exchange for personal travel time gains. However, many scholars have argued that such travel choice decisions of individuals in their role of consumer of mobility are likely to be a poor proxy of how they in their role of citizen believe government should spend tax money to generate travel time gains for large numbers of travelers. So far, this possible deviation between what we call ‘consumer VoT’ and ‘citizen VoT’ has not been studied empirically. In this paper, we fill this gap, by designing a Stated Choice experiment with eight different frames; some representing a typical consumer choice situation, others gradually approaching a citizen perspective. We find that individuals’ willingness to pay from previously collected tax money for travel time gains created by a government policy, is significantly higher than their willingness to pay, from their after tax income, for time gains obtained by choosing a different route. This result implies that citizen VoT is higher than consumer VoT. This difference does not stem from a stronger willingness to spend previously collected tax money compared to spending one’s own income, but from a difference in the value attached to travel gains: a travel time gain resulting from government action is valued more than the same travel time gain obtained by one’s own route choices. This and a range of other empirical results are discussed in depth, in light of the conceptual differences between preferences of individuals in a role of consumer or citizen.