- Home
- Research
- Core Topics
- Transportation Network Modeling and Planning
Transportation Network Modeling and Planning
Highway Managed Lane Usage and Tolling for Mixed Traffic Flows with Connected Automated Vehicles and High-Occupancy Vehicles
Researcher(s): Max Ng, Hani Mahmassani
Year: 2023
This paper investigates managed lane toll setting and its effect under mixed traffic of connected automated vehicles (CAVs), high-occupancy vehicles (HOVs), and human-driven vehicles (HDVs), with the goal of avoiding flow breakdown and minimizing total social cost. A mesoscopic finite difference traffic simulation model considers the flow–density relationship at different CAV market penetration rates, lane-changing behaviors, and multiple entries/exits, interacting with a reactive toll setting mechanism. The results of Monte Carlo simulation suggest an optimal policy of untolled HOV/CAV use with tolled HDVs in particular scenarios of limited CAV market penetration. Small and targeted tolling avoids flow breakdown in managed lanes while prioritizing HOVs and other vehicles with high values of time. Extensions of the formulation and sensitivity analysis quantify the benefits of converting high-occupancy HDVs to CAVs. The optimal tolling regime combines traffic science notions of flow stability and the economics of resource allocation.
Potential of carpool for network traffic management
Year: 2022
This study examines the impact of carpool on network traffic in a highly idealized futuristic world, where all travelers are willing to participate in carpool arranged by a Transportation Network Company. We build a parsimonious carpool model that focuses on the trade-off between inconvenience costs and travel cost savings. Underlying the model is a nonlinear bipartite matching problem that seeks to maximize commuters’ welfare. By assuming the congestion effect is negligible, we derive several useful analytical results. When the inconvenience cost is less than the median trip valuation of a rider, the platform could always achieve an almost perfect match while maximizing commuters’ welfare, which corresponds to a 50% reduction in vehicular traffic flow. In the case of perfect match, if there is an even number of travelers, we propose a pricing policy that possesses all desired properties of the Vickrey-Clark-Groves (VCG) policy – a benchmark truthful policy for achieving socially optimal solution – but runs a lower deficit. Otherwise, we show the VCG policy always generates a profit. If the inconvenience cost is too high, the perfect match is no longer socially optimal, but the VCG policy still yields a positive profit. Solutions from numerical experiments generally agree with the analytical results. They also suggest that matching across O-D pairs occurs only when it has a significantly lower inconvenience cost than matching within, an unlikely event in reality. Moreover, when cross O-D matching does become prevalent, it leads to higher vehicle miles travelled, hence worse congestion. Thus, from the point of view of traffic management, cross O-D carpool should not be encouraged.
Read the full publication
Understanding Ridesplitting Behavior with Interpretable Machine Learning Models Using Chicago Transportation Network Company Data
Year: 2022
As congestion levels increase in cities, it is important to analyze people’s choices of different services provided by transportation network companies (TNCs). Using machine learning techniques in conjunction with large TNC data, this paper focuses on uncovering complex relationships underlying ridesplitting market share. A real-world dataset provided by TNCs in Chicago is used in analyzing ridesourcing trips from November 2018 to December 2019 to understand trends in the city. Aggregated origin–destination trip-level characteristics, such as mean cost, mean time, and travel time reliability, are extracted and combined with origin–destination community-level characteristics. Three tree-based algorithms are then utilized to model the market share of ridesplitting trips. The most significant factors are extracted as well as their marginal effect on ridesplitting behavior, using partial dependency plots for interpretation of the machine learning model results. The results suggest that, overall, community-level factors are as or more important than trip-level characteristics. Additionally, the percentage of White people highly affects ridesplitting market share as well as the percentage of bachelor’s degree holders and households with two people residing in them. Travel time reliability and cost variability are also deemed more important than travel time and cost savings. Finally, the potential impact of taxes, crimes, cultural differences, and comfort is discussed in driving the market share, and suggestions are presented for future research and data collection attempts.
Read the full publication
Are autonomous vehicles better off without signals at intersections? A comparative computational study
Year: 2022
We model and analyze a futuristic intersection that serves only connected, autonomous and centrally managed vehicles. Under consideration are three control strategies that aim to minimize the total system delay by choosing an optimal trajectory for each vehicle. The first two abandon the concept of signal timing all together whereas the third strategy keeps it. The difference between the two signal-free strategies has to do with a fail-safe buffer requirement introduced to provide redundancy. Each control strategy leads to a unique version of a trajectory-based autonomous intersection management (T-AIM) problem, which is formulated as a mixed integer linear program and solved using both a commercial solver and a specialized heuristic algorithm. We find the signal-free strategy holds an overwhelming advantage over the signal-based strategy in terms of efficiency. However, its success is fragile and dependent on the faith in the safety and reliability of the system. When the fail-safe buffer is introduced, the efficiency of the signal-free strategy degrades to a level comparable to that of a properly optimized signal-based strategy. Surprisingly, the signal-free strategy with redundancy tends to arrange vehicles in groups that take turns to cross the intersection together. This “signal-like behavior” manifests itself whenever congestion rises to a certain threshold. In addition, solving the T-AIM problem based on signal timing enjoys significant computational benefits, because it eliminates many conflicts. Thus, the basic logic of signal timing – if not the physical equipment – may survive even after humans are no longer allowed to drive.
Read the full publication
Does intercity rail station placement matter? Expansion of the node-place model to identify station location impacts on Amtrak ridership
Year: 2022
The node-place model has been used in previous studies to categorize urban transit rail stations, and to study impacts on transit station ridership. Similar studies have not been performed for intercity rail station ridership. This study uses the node-place model to examine the station-level factors affecting station ridership on the Amtrak network in the United States. The local factors include measures of the node and place quality of each station. The node-place model is expanded to include accessibility, an important consideration for the large catchment areas of intercity rail. Measures and indices defining each node-place-accessibility category are constructed and analyzed in two groups of Amtrak stations. A multivariate regression is used to determine the effects of each measure and category on the station ridership. The results indicate that the quality of place and accessibility for stations significantly impact station ridership for both groups of Amtrak stations. The findings contribute to a better understanding of the drivers of intercity rail ridership; the resulting insights could be used to improve service and development planning for intercity rail networks.
Read the full publication
Interdependence in active mobility adoption: Joint modeling and motivational spillover in walking, cycling and bike-sharing
Year: 2022
Active mobility offers an array of physical, emotional, and social well-being benefits. However, with the proliferation of the sharing economy, new nonmotorized means of transport are entering the fold, complementing some existing mobility options while competing with others. The purpose of this research study is to investigate the adoption of three active travel modes—namely walking, cycling, and bikesharing—in a joint modeling framework. The analysis is based on an adaptation of the stages of change framework, which originates from the health behavior sciences. Multivariate ordered probit modeling drawing on U.S. survey data provides well-needed insights into individuals’ preparedness to adopt multiple active modes as a function of personal, neighborhood, and psychosocial factors. The research suggests three important findings. (1) The joint model structure confirms interdependence among different active mobility choices. The strongest complementarity is found for walking and cycling adoption. (2) Each mode has a distinctive adoption path with either three or four separate stages. We discuss the implications of derived stage-thresholds and plot adoption contours for selected scenarios. (3) Psychological and neighborhood variables generate more coupling among active modes than individual and household factors. Specifically, identifying strongly with active mobility aspirations, experiences with multimodal travel, possessing better navigational skills, along with supportive local community norms are the factors that appear to drive the joint adoption decisions. This study contributes to the understanding of how decisions within the same functional domain are related and help to design policies that promote active mobility by identifying positive spillovers and joint determinants.
Read the full publication
Reliable trajectory-adaptive routing strategies in stochastic, time-varying networks with generalized correlations
Researcher(s): Monika Filipovska, Hani Mahmassani
Year: 2021
This paper focuses on the problem of finding optimal trajectory-adaptive routing strategies in stochastic time-varying networks with generalized spatio-temporal correlations. A representation for jointly distributed continuous link travel times across the entire network with time-varying distributions and correlation structures is presented, and the crucial characteristics and methodological difficulties of the problem are discussed. The paper presents a generalized 2-stage path and strategy finding solution approach that can serve for finding both exact and approximate solutions with the tuning of a risk-level tolerance parameter. The first stage of the solution approach generates eligible paths, where the risk-level parameter is used to eliminate paths that are likely to be inefficient. The second stage finds reliable trajectory-adaptive strategies, using the eligible paths only, based on one or multiple reliability-based optimality conditions. Thus, the approach allows the user to determine the optimal strategy for one or multiple groups of travelers with different reliability preferences. Numerical experiments show that the average running time of the algorithm reduces super-linearly with the increase of the risk-tolerance parameter ∊, while incurring some loss to the objective function relative to the exact solution. Thus, the heuristic can offer significant benefits in reducing the run time of the solution algorithm, while finding adaptive strategy solutions that consistently maintain better objective function values compared to the a priori (i.e., non-adaptive) solutions.
Macroscopic network-level traffic models: Bridging fifty years of development toward the next era
Year: 2021
Network macroscopic fundamental diagrams (NMFD) and related network-level traffic dynamics models have received both theoretical support and empirical validation with the emergence of new data collection technologies. However, the extent to which network-level macroscopic traffic models may be ready for practical implementation remains to be ascertained. This paper aims to shed light on this matter by reviewing the 50-year history of macroscopic modeling of urban networks and assessing remaining gaps and opportunities for further development of both theory and applications. To this end, the existing literature's chronology is mapped onto three eras of development, and the corresponding theories, assumptions, and limitations are outlined and discussed in two streams, equilibrium relations and traffic dynamics. Among the topics pertaining to equilibrium relations, the highlighted gaps include the lack of empirical studies on the hysteresis and bifurcation phenomena, the existence of multi-modal NMFD (3D-NMFD) in different traffic conditions, the factors that might affect the (3D-) NMFD shape, the accuracy of speed-NMFDs in particular different bus speed NMFDs, and the passenger-oriented NMFDs. Research gaps pertaining to traffic dynamics include the analytical solution of trip-based models, the FIFO violation in the delay-based models, the definition of outflow and entrance functions, the notion of active network length, the trip length distribution, and the path flow distribution. Future research directions targeting topics that might shape the potential next era include the practical implementation of NMFD-based control strategies, the application of NMFD in quality of service assessments, and NMFD in the presence of new technologies such as connected and autonomous vehicles (CAVs).
Read the full publication
Reliable trajectory-adaptive routing strategies in stochastic, time-varying networks with generalized correlations
Year: 2021
This paper focuses on the problem of finding optimal trajectory-adaptive routing strategies in stochastic time-varying networks with generalized spatio-temporal correlations. A representation for jointly distributed continuous link travel times across the entire network with time-varying distributions and correlation structures is presented, and the crucial characteristics and methodological difficulties of the problem are discussed. The paper presents a generalized 2-stage path and strategy finding solution approach that can serve for finding both exact and approximate solutions with the tuning of a risk-level tolerance parameter. The first stage of the solution approach generates eligible paths, where the risk-level parameter is used to eliminate paths that are likely to be inefficient. The second stage finds reliable trajectory-adaptive strategies, using the eligible paths only, based on one or multiple reliability-based optimality conditions. Thus, the approach allows the user to determine the optimal strategy for one or multiple groups of travelers with different reliability preferences. Numerical experiments show that the average running time of the algorithm reduces super-linearly with the increase of the risk-tolerance parameter ∊, while incurring some loss to the objective function relative to the exact solution. Thus, the heuristic can offer significant benefits in reducing the run time of the solution algorithm, while finding adaptive strategy solutions that consistently maintain better objective function values compared to the a priori (i.e., non-adaptive) solutions.
Read the full publication
Estimation of path travel time distributions in stochastic time-varying networks with correlations
Year: 2021
Transportation research has increasingly focused on the modeling of travel time uncertainty in transportation networks. From a user’s perspective, the performance of the network is experienced at the level of a path, and, as such, knowledge of variability of travel times along paths contemplated by the user is necessary. This paper focuses on developing approaches for the estimation of path travel time distributions in stochastic time-varying networks so as to capture generalized correlations between link travel times. Specifically, the goal is to develop methods to estimate path travel time distributions for any path in the networks by synthesizing available trajectory data from various portions of the path, and this paper addresses that problem in a two-fold manner. Firstly, a Monte Carlo simulation (MCS)-based approach is presented for the convolution of time-varying random variables with general correlation structures and distribution shapes. Secondly, a combinatorial datamining approach is developed, which aims to utilize sparse trajectory data for the estimation of path travel time distributions by implicitly capturing the complex correlation structure in the network travel times. Numerical results indicate that the MCS approach allowing for time-dependence and a time-varying correlation structure outperforms other approaches, and that its performance is robust with respect to different path travel time distributions. Additionally, using the path segmentations from the segment search approach with a MCS approach with time-dependence also produces accurate and robust estimates of the path travel time distributions with the added benefit of shorter computation times.
Read the full publication
Developing a Merge Lane Change Decision Policy for Autonomous Vehicles by Deep Reinforcement Learning
Year: 2021
With autonomous vehicles (AVs) being actively developed, it becomes possible to optimize vehicle control policies and traffic management tools in a mixed vehicular environment. For individual AV control, acceleration and lane change are the two elementary driving behaviors that need to be coordinated to minimize disturbance to the entire traffic dynamics. In this paper, a joint decision policy of acceleration and lane change actions for AVs on a merging ramp is proposed and trained in a mixed autonomy traffic, using the technique of deep reinforcement learning. Our method is able to train policies that have limited impact on highway traffic while maintaining a relatively high merge throughput. We experimented with two reward functions, designed for the AV's selfish benefits and for the network traffic's social benefits. This paper then examines the emergent behaviors exhibited by the trained policies and their impacts on the main highway traffic at different density levels.
Read the full publication
Estimating network travel time reliability with network partitioning
Year: 2020
Network travel time reliability can be represented by a relationship between network space-mean travel time and the standard deviation of network travel time. The primary objective of this paper is to improve estimation of the network travel time reliability with network partitioning. We partition a heterogeneous large-scale network into homogeneous regions (clusters) with well-defined Network Fundamental Diagrams (NFD) using directional and non-directional partitioning approaches. To estimate the network travel time reliability, a linear relationship is estimated that relates the mean travel time with the standard deviation of travel time per unit of distance at the network level. The impact of different partitioning approaches, as well as the number of clusters, on the network travel time reliability relationship are also explored. To estimate individual vehicle travel times, we use two distinct approaches to allocate vehicle trajectories to different time intervals, namely trajectory and sub-trajectory methods. We apply the proposed framework to a large-scale network of Chicago using a 24-h dynamic traffic simulation. Partitioning and travel time reliability estimation are conducted for both morning and afternoon peak periods to demonstrate the impacts of travel demand pattern variations. The numerical results show that the sub-trajectory method for the network travel time reliability estimation and the directional partitioning with three clusters have the highest performance among other tested methods. The analyses also demonstrate that partitioning a heterogeneous network into homogeneous clusters may improve network travel reliability estimation by estimating an independent relationship for each cluster. Also, comparing morning and afternoon peak periods suggests that the estimated parameter for the linear network travel time reliability relationship is directly related to the coefficient of variation of density as a measure of spatial distribution of congestion across the network.
Read the full publication
Joint design of multimodal transit networks and shared autonomous mobility fleets
Year: 2020
Providing quality transit service to travelers in low-density areas, particularly travelers without personal vehicles, is a constant challenge for transit agencies. The advent of fully-autonomous vehicles (AVs) and their inclusion in mobility service fleets may allow transit agencies to offer better service and/or reduce their own capital and operational costs. This study focuses on the problem of allocating resources between transit patterns and operating (or subsidizing) shared-use AV mobility services (SAMSs) in a large metropolitan area. To address this question, a joint transit network redesign and SAMS fleet size determination problem (JTNR-SFSDP) is introduced, and a bi-level mathematical programming formulation and solution approach are presented. The upper-level problem modifies a transit network frequency setting problem (TNFSP) formulation via incorporating SAMS fleet size as a decision variable and allowing the removal of bus routes. The lower-level problem consists of a dynamic combined mode choice-traveler assignment problem (DCMC-TAP) formulation. The heuristic solution procedure involves solving the upper-level problem using a nonlinear programming solver and solving the lower-level problem using an iterative agent-based assignment-simulation approach. To illustrate the effectiveness of the modeling framework, this study uses traveler demand from Chicago along with the region's existing multimodal transit network. The computational results indicate significant traveler benefits, in terms of improved average traveler wait times, associated with optimizing the joint design of multimodal transit networks and SAMS fleets compared with the initial transit network design.
Read the full publication
Real-Time Traffic Flow Pattern Matching to Improve Predictive Performance of Online Simulation-Based Dynamic Traffic Assignment
Year: 2019
In real-time simulation-based dynamic traffic assignment, selection of the most suitable demand from the library of demands calibrated offline improves the accuracy of the prediction. In the era of data explosion, relying on contextual and rule-based pattern matching logic does not seem sufficient. A rolling horizon scheme for real-time pattern matching is introduced using two pattern matching frameworks. The hard matching algorithm chooses the closest pattern at each evaluation interval, while soft matching calculates the probability of being a match for each pattern. To make sure the pattern switch does not happen because of short-lived interruptions in traffic conditions, a persistency index is introduced. The results show that the number of switches in hard matching is bigger than soft matching but the error of real-time matching for both cases is low. The importance of the results is twofold: First, any observation that is not similar to only one pattern in the library can be mimicked using multiple available patterns; second, more advanced algorithms can match the patterns existing in the library, without any contextual logics for pattern matching.
Read the full publication
Privately Owned Autonomous Vehicle Optimization Model Development and Integration with Activity-Based Modeling and Dynamic Traffic Assignment Framework
Year: 2019
This paper presents a first-order approach integrated with activity-based modeling and dynamic traffic assignment framework to model the impact of autonomous vehicles on household travel and activity schedules. By considering shared rides among household members, mode choices, re-planning of departure times, and the rescheduling of activity sequences, two optimization models—basic personal owned autonomous vehicle (POAV) model and enhanced POAV model—are presented. The proposed approach is tested for the different models at the household level with different household sizes. The activity schedules of each household were generated in the Chicago sub-area network. The results show that each POAV can effectively replace multiple conventional vehicles, however, using POAV will lead to more vehicle miles traveled because of detour trips. The proposed enhanced POAV model considers mode choice decision with a household-based approach instead of a trip-based approach to capture the impacts of repositioning trips on mode choice. The results show that, if the generalized travel cost of POAV remains at the same level as conventional vehicles, more passengers will choose to use transit because the repositioning trips increase the total cost.
Read the full publication
Day-to-Day Learning Framework for Online Origin–Destination Demand Estimation and Network State Prediction
Year: 2019
Origin–destination (O–D) demand is a critical component in both online and offline dynamic traffic assignment (DTA) systems. Recent advances in real-time DTA applications in large networks call for robust and efficient methodologies for online O–D demand estimation and prediction. This study presents a day-to-day learning framework for a priori O–D demand, along with a predictive data-driven O–D correction approach for online consistency between predicted and observed (sensor) values. When deviations between simulation and real world are observed, a consistency-checking module initiates O–D demand correction for the given prediction horizon. Two predictive correction methods are suggested: 1) simple gradient method, and 2) Taylor approximation method. New O–D demand matrices, corrected for 24 simulation hours by the correction module, are used as the updated a priori demand for the next day simulation. The methodology is tested in a real-world network, Kansas City, MO, for a 3-day period. Actual tests in real-world networks of online DTA systems have been very limited in the literature and in actual practice. The test results are analyzed in time and space dimensions. The overall performance of observed links is assessed. To measure the impact of O–D correction and daily O–D updates, traffic prediction performance with the new modules is compared with the base case. Predictive O–D correction improves prediction performance in a long prediction window. Also, daily updated O–D demand provides better initial states for traffic prediction, enhancing prediction in short prediction windows. The two modules collectively improve traffic prediction performance of the real-time DTA system.
Read the full publication
Operational Scenario Definition in Traffic Simulation-Based Decision Support Systems: P
Year: 2019
This paper is intended to mine historical data by presenting a scenario clustering approach to identify appropriate scenarios for mesoscopic simulation as a part of the evaluation of transportation projects or operational measures. It provides a systematic and efficient approach to select and prepare effective input scenarios for a given traffic simulation model. The scenario clustering procedure has two primary applications: travel time reliability analysis, and traffic estimation and prediction systems. The ability to systematically identify similarity and dissimilarity among weather scenarios can facilitate the selection of critical scenarios for reliability studies. It can also support real-time weather-responsive traffic management (WRTM) by quickly classifying a current or predicted weather condition into predefined categories and suggesting relevant WRTM strategies that can be tested via real-time traffic simulation before deployment. A detailed method for clustering weather time series data is presented and demonstrated using historical data. Two clustering algorithms with different similarity measures are compared. Clustering results using a k-means clustering algorithm with squared Euclidean distance are illustrated in the travel time reliability application.
Read the full publication
Prediction and Mitigation of Flow Breakdown Occurrence for Weather Affected Networks: Case Study of Chicago, Illinois
Year: 2019
This study investigates the prediction and mitigation of the phenomenon of traffic flow breakdown when affected by varying weather conditions. First, the probability of breakdown occurrence is examined using a survival analysis approach to obtain distributions of pre-breakdown flow rates under different weather conditions. Second, pre-breakdown flow rate distributions were applied in breakdown prediction for the implementation of breakdown mitigation strategies. In the first part, a set of data from the network of Kansas City was used to demonstrate the applicability of the Kaplan–Meier Product Limit method to estimating the breakdown probability under various weather conditions. Then, using simulated data on the network of Chicago, the K-M approach was used again to obtain survival likelihood distributions, which in turn yield breakdown probability, for 13 different weather cases as combinations of weather categories for different levels of visibility, rain, and snow precipitation. In the second part, continuing with the simulated data, dynamic speed limits (DSL) were applied to demonstrate the effectiveness of the prediction method presented. A sensitivity analysis of the threshold probability and upstream distance at which DSL should be implemented was performed for clear and inclement weather conditions. In clear weather the performance of the strategy is better at a lower probability threshold and farther upstream location, whereas in inclement weather the performance is better at a lower probability threshold and closer upstream location. The paper demonstrates the effect of changing weather conditions on the likelihood of breakdown occurrence and the implementation of breakdown mitigation strategies.
Read the full publication
Network Positioning from the Edge: An Empirical Study of the Effectiveness of Network Positioning in P2P Systems
Researcher(s): David R. Choffnes, Mario A. Sanchez, Fabian E. Bustamante
Year: 2010
Network positioning systems provide an important service to large-scale P2P systems, potentially enabling clients to achieve higher performance, reduce cross-ISP traffic and improve the robustness of the system to failures. Because traces representative of this environment are generally unavailable, and there is no platform suited for experimentation at the appropriate scale, network positioning systems have been commonly implemented and evaluated in simulation and on research testbeds. The performance of network positioning remains an open question for large deployments at the edges of the network.
This paper evaluates how four key classes of network positioning systems fare when deployed at scale and measured in P2P systems where they are used. Using 2 billion network measurements gathered from more than 43,000 IP addresses probing over 8 million other IPs worldwide, we show that network positioning exhibits noticeably worse performance than previously reported in studies conducted on research testbeds. To explain this result, we identify several key properties of this environment that call into question fundamental assumptions driving network positioning research.
Relationship between Proximity to Transit and Ridership for Journey-to-Work Trips in Chicago
Researcher(s): Lindsey Marshall, Joseph L. Schofer, Pablo Durango-Cohen, Kimberly A. Gray
Year: 2010
This circular summarizes discussions at a peer exchange of state department of transportation officials and other professionals that focused on data and information uses, management strategies, needs, and gaps in their organizations. The peer exchange examined the role of data and information in transportation decision making; identified information resources, gaps, and opportunities; and explored data, access, and analysis improvements for information resource programs. In addition, participants discusses possible strategies that the transportation community might use to implement such improvements.
Stability of User-equilibrium Route Flow Solutions for the Traffic Assignment Problem
Researcher(s): Shu Lu, Yu Nie
Year: 2010
This paper studies stability of user-equilibrium (UE) route flow solutions with respect to inputs to a traffic assignment problem, namely the travel demand and parameters in the link cost function. It shows, under certain continuity and strict monotonicity assumptions on the link cost function, that the UE link flow is a continuous function of the inputs, that the set of UE route flows is a continuous multifunction of the inputs, and that the UE route flow selected to maximize an objective function with certain properties is a continuous function of the inputs. The maximum entropy UE route flow is an example of the last. On the other hand, a UE route flow arbitrarily generated in a standard traffic assignment procedure may not bear such continuity property, as demonstrated by an example in this paper.
Toward More Reliable Mobility: Improved Decision Support Tools for Transportation Systems
Researcher(s): Yu Nie
Year: 2010
The overarching goal of the project is to enhance travel reliability of highway users by providing them with reliable route guidance produced from newly developed routing algorithms that are validated and implemented with real traffic data. Phase I of the project (funded by CCITT in 2008) focused on demonstrating the value of reliable route guidance through the development of dissemination of Chicago Testbed for Reliable Routing (CTR). Phase II aims at bringing the implementation of reliable-routing technology to the next stage through initial deployment of CTR.
The first objective in Phase II is to create a travel reliability inventory (TRI) of Northeastern Illinois using CTR, by collaborating with public agencies such as the Illinois Department of Transportation (IDOT), Chicago Transit Authority (CTA) and Chicago Traffic Management Authority (CTMA). TRI documents travel reliability indices (e.g., 95 percentile route travel times) between heavily-traveled origins-destination pairs in the region, which are of interest not only to individual travel decision-making, but also regional transportation planning and traffic operations/management. The second objective is to perform and initial market test in order to understand users’ need for and response to reliability information and reliable route guidance.
To these ends, the following research activities are proposed to further develop CTR:
- Implement and test latest reliable routing algorithms that are suitable for large-scale applications.
- Develop a web-based version of CTR and host the service at Northwestern University’s Translab Website. A web survey will be designed and posted along with CTR in order to collect user feedback.
- Explore the possibility of achieving a greater degree of data coverage of the study area. Specifically, archived automatic vehicle location (AVL) data of CTA’s bus fleet is considered an important data source to supplement GCM data and will receive a focal study.
Crowdsourcing Service-level Network Event Monitoring
Researcher(s): David R. Choffnes, Fabián E. Bustamante, Zihui Ge
Year: 2010
The user experience for networked applications is becoming a key benchmark for customers and network providers. Perceived user experience is largely determined by the frequency, duration and severity of network events that impact a service. While today’s networks implement sophisticated infrastructure that issues alarms for most failures, there remains a class of silent outages (e.g., caused by configuration errors) that are not detected. Further, existing alarms provide little information to help operators understand the impact of network events on services. Attempts to address this through infrastructure that monitors end-to-end performance for customers have been hampered by the cost of deployment and by the volume of data generated by these solutions. We present an alternative approach that pushes monitoring to applications on end systems and uses their collective view to detect network events and their impact on services - an approach we call Crowdsourcing Event Monitoring (CEM). This paper presents a general framework for CEM systems and demonstrates its effectiveness for a P2P application using a large dataset gathered from BitTorrent users and confirmed network events from two ISPs. We discuss how we designed and deployed a prototype CEM implementation as an extension to BitTorrent. This system performs online service-level network event detection through passive monitoring and correlation of performance in end-users’ applications.
Field Test of a Method for Finding Consistent Route Flows and Multiple-Class Link Flows in Road Traffic Assignments
Researcher(s): David Boyce,
Yu (Marco) Nie, Hillel Bar-Gera, Yang Liu, and Yucong Hu
Year: 2010
Road traffic assignment, or forecasting route and link flows corresponding to fixed matrices of origin-destination (OD) flows by vehicle class on a road network for a given time period, is commonly applied by transportation planning practitioners. The standard user-equilibrium traffic assignment method uniquely determines the total flow on each network link, subject to convergence errors. Multiple-class link flows and route flows, however, are indeterminate. To ensure that route and multiple-class link flows are uniquely determined, or consistent, an additional assumption is required. One option is that proportions of flow over alternative route segments with equal costs are the same for all drivers, regardless of origin or destination. Analyses based on the assigned link and route flows by vehicle class, such as select link, select zone and emissions analyses, are often performed without considering this issue. Although such analyses have become important in practice, no commercial software system currently considers the indeterminacy of these flows.
Traffic Assignment by Paired Alternative Segments (TAPAS) is a new algorithm offering the first practical way to address this issue. In this project six practitioners analyzed how route flows and/or multiple-class link flows generated by TAPAS compared with those found by the commercial software systems. A specialized tool VPAS was developed to compare the outputs of TAPAS and the practitioner software. The project team also undertook its own case study of the Chicago region with tools offered by four commercial software systems, which may be classified into two groups: link-based and quick-precision. Link-based tools applied in the project were CUBE, EMME, and TransCAD; quick precision tools applied were VISUM’s route-based method and TransCAD’s origin user-equilibrium (OUE) method. Findings of these applications may be summarized as follows:
- Select link results for link-based tools were approximately similar to those found by TAPAS; differences in flows through a selected link by OD pair were relatively small. However, small flows were observed in link-based solutions on non-equilibrium routes not found in the more precise TAPAS solutions. As a result, the number of OD pairs using a select link was often much larger for link-based tools than for TAPAS. Analyses of flows on pairs of equal-cost segments showed that link-based solutions tended to satisfy approximately the proportionality condition. Slow convergence, however, is a costly limitation of link-based tools. Even so, the findings suggest that link-based tools do provide approximately proportional solutions, which was not realized before this project.
- Select link results for quick-precision tools were very different from those produced by TAPAS. In particular, where TAPAS predicted positive flows, quick-precision tools often gave zero flow from an OD pair through a selected link. Analyses of flows on pairs of equal-cost segments showed that quick-precision tools produced solutions that violate the proportionality condition. In two-class assignments for pairs of alternative segments, the proportions of flow found by quick-precision solutions were also different by class.
Improving Our Understanding of How Pricing and Congestion Affect Travel Demand
Researcher(s): Hani Mahmassani
Year: 2010
Dr. Mahmassani serves as co-PI on this study, awarded by the National Academy of Science to PB Americas, Inc. (with Northwestern University Transportation Center). The work plan can be conceptualized in three interconnected levels of behavioral rigor and practical application, with varying levels of sophistication:
Level 1 – Behavioral Foundations. The first level corresponds to behavioral models intended for a deep understanding and quantitative exploration of travel behavior. These models seek to address the full range of possible short and long‐term responses, but also may focus on select choice dimensions (for example, route and departure time choices, or usual workplace location choice).
Level 2 – Advanced Operational. The second level relates to relatively advanced, yet operational, Activity‐Based (AB) models, integrated with state of the art DTA (Dynamic Traffic Assignment) models. These models allows for the incorporation of a wide range of possible short‐ and long‐term responses that are embedded in the choice hierarchy. The integrity of operational models requires that each and every choice dimension should be allocated a proper “slot” in the hierarchy, with upward and downward linkages to related choices. Operational/computing time requirements often limit the total number of choice dimensions and alternatives, but this restriction is lessening with time.
Level 3 – Opportunities for Prevailing Practice. The third level relates to existing model systems used by most of MPOs and state DOTs, in the form of aggregate trip-based models (frequently referred to as 4-step models). Though rather restrictive in design, such models offer opportunities for meaningful and immediate contributions to the state of travel demand modeling practice. A serious restriction of 4‐step models is that these rely on static assignment procedures. Static assignments generate only crude average travel time and cost variables, and reliability can be incorporated only through certain proxies.
The SHRP 2 C04 project has completed an inventory of available datasets to support the research, and demonstrated an integrated application of user response models with a simulation-based DTA platform for the New York region Best Practice Model network.
Network Design for Code Sharing
Researcher(s): Diego Klabjan
Year: 2010
An airline from an alliance faces the daunting task of code sharing its flights. The challenge mainly lies in the sheer size of the itineraries that can be sold on the entire network of all alliance partners. We developed a network design approach based on discrete choice modeling of passengers' utilities. The solution recommends flights to code share. In comparison to existing designs, our solution attains up to 2% improved profit, which was evaluated by a commercial profitability model.
Incorporating Reliability Performance Measures in Operations and Planning Modeling Tools
Researcher(s): Hani Mahmassani
Year: 2009
Dr. Mahmassani is one of three principal investigators on the team led by Delcan, Inc. to undertake this project. Northwestern’s role focuses on the theoretical and methodological underpinnings of integrated supply-demand models that incorporate reliability. The objectives are to advance the state of the art in planning and operations models to produce measures of reliability performance of proposed system changes, and determine how travel demand forecasting models can use reliability measures to produce more realistic estimates of travel patterns. Project L04 draws on the quantitative measures of reliability as well as the impacts of reliability on route choice, time-of-day choice, and mode choice substantiated in “Improving Our Understanding of How Highway Congestion and Pricing Affect Travel Demand, SHRP2 C04”.
This project is developing approaches and tools to incorporate reliability as an input as well a key output in traffic models used for both operations and planning applications. A unifying framework for reliability analysis is proposed, applicable in conjunction with any particle-based micro- or meso- simulation model that produces trajectories. Vehicle trajectories are introduced and discussed as a central building block in this framework. The methodology is demonstrated using a simulation-based DTA platform.
In addition, to capture travel time variability introduced by random events, a repeatable framework is developed for modeling and evaluating incidents and events. A key variability-inducing phenomenon is traffic flow breakdown, which is modeled as an inherently stochastic phenomenon with structural dependence on state variables of the system. Reliability-improving measures highlighted in the report include information supply and dynamic pricing, whose effectiveness increases considerably when applied in real-time on the basis of predicted conditions.
Finally, possible applications of travel time reliability in operations-oriented models are presented.
Incorporating Weather Impacts in Traffic Estimation and Prediction Systems
Researcher(s): Hani Mahmassani
Year: 2009
Dr. Mahmassani served as PI for this study conducted for FHWA under a subcontract to SAIC, Inc. The objectives of the project are to develop weather-sensitive traffic prediction and estimation models and incorporate them in existing traffic estimation and prediction systems. This includes enhancement of the capabilities in mesoscopic DTA tools to model traffic behavior under inclement weather, and capture user responses to inclement weather with and without the presence of advisory and control strategies.
As a result of this project, The DYNASMART TrEPS can now capture the effects of adverse weather on traffic patterns through both supply and demand side modifications to the model. New weather‐related features in DYNASMART include:
Weather Scenario Specification: DYNASMART allows users to specify various weather scenarios for the study network. It can be represented as either the network-wide weather condition or the link‐specific weather condition.
Weather Adjustment Factor: Users can define the effect of weather on supply‐side traffic parameters such as free flow speed and capacity based on three weather condition parameters: visibility (mile), rain precipitation intensity (inch/hr) and snow precipitation intensity (inch/hr) by means of Weather Adjustment Factors (WAF). DYNASMART applies user‐specified WAF to 18 supply‐side traffic properties for links within the impacted region to simulate traffic conditions under the weather condition. WAF can be obtained based on calibrated weather‐traffic flow relation.
Modeling Traffic Advisory and Control via Variable Message Signs (VMS): DYNASMART provides three weather‐related VMS operation functionalities: (1) Speed Reduction Warning – via a VMS warning sign indicating low visibility (e.g., fog) or slippery road (e.g. rain and snow), speed reduction behavior under adverse weather can be simulated; (2) Optional Detour – VMS suggests that travelers re-evaluate their current route based on the generalized cost that includes travel penalties of the added delays caused by adverse weather; and (3) Variable Speed Limit (VSL) – in DYNASMART, vehicle speed can be regulated through the speed limits posted on VMS in correspondence with prevailing weather conditions.
Optimal Short-Range Routing of Vessels in a Seaway
Researcher(s): I.S. Dolinskaya, M. Kotinis, M.G. Parsons, R.L. Smith, R. L
Year: 2009
An investigation of the optimal short-range routing of a vessel in a stationary random seaway is presented. The calculations are performed not only in head seas but also in oblique waves. The evaluation of the added drag is performed by computing the time average wave force acting on the vessel in the longitudinal direction. Subsequently, the added drag is superimposed on the steady drag experienced by the ship as it advances in calm water. In this manner, the fastest path between the origin point A and the destination point B can be evaluated, taking into account operational constraints. To obtain the fastest path between two points, the underlying structure and properties of the maximum mean attainable speed are analyzed. This detailed analysis allows us to demonstrate the fastest path for the problem without any operational constraints to be a straight line. Subsequently, the solution is reevaluated for scenarios where the original optimal path violates at least one of the operability criteria considered. For that case, a fastest path is found to be a path consisting of one waypoint, that is, a two line segment path. In addition to providing a closed-form fastest-path solution for the case of no operational constraints, a bound is provided for travel time error for more general speed functions in the case where a straight line path is followed.
Providing Reliable Route Guidance Using Chicago Data
Researcher(s): Yu (Marco) Nie, Xing Wu
Year: 2009
New techniques offer the potential to improve travel reliability for motorists, freight carriers and parcel delivery firms. This project confronts challenges to the implementation of these techniques, and demonstrated their feasibility and benefits using real data from the Chicago metropolitan area, one of the largest transportation hubs in the US. Conceptually, the most reliable routes can be found by solving the Dynamic Shortest Path problem with On-Time arrival reliability (DSPOT). DSPOT has recently been formulated and solved using the dynamic programming technique. The proposed research addresses two important issues that currently preclude its implementation: 1) development of solution algorithms fast enough for on-line application, and 2) validation using real data. In this project, historical traffic data from the Gary-Chicago-Milwaukee 9GCM) traveler information system will be used to prepare dynamic probability mass functions of travel times, which are the key inputs to DSPOT. Then a prototype path search tool will be developed, which implements DSPOT based on GCM data. This toll will be made available to the public through the Artificial Intelligence Laboratory at the University of Illinois at Chicago. The ultimate goal of this project is to provide motorists and carrier with commercialized DSPOT products that will allow them to make tradeoffs between reliability and other costs and constraints. With the benefits and market value demonstrated through this project and further implementation stages, we believe that the related industries will be interested in adding DSPOT to their product offerings. These firms include but are not limited to the manufacturers of in-vehicle navigation systems, web companies that provide internet-based driving directions and software vendors that produce logistics for freight carriers.
Anomalous Diffusion and the Structure of Human Transportation Networks
Researcher(s): Dirk Brockmann
Year: 2008
The dispersal of individuals of a species is the key driving force of various spatiotemporal phenomena which occur on geographical scales. It can synchronise populations of interacting species, stabilise them, and diversify gene pools [1–3]. The geographic spread of human infectious diseases such as influenza, measles and the recent severe acute respiratory syndrome (SARS) is essentially promoted by human travel which occurs on many length scales and is sustained by a variety of means of transportation [4–8]. In the light of increasing international trade, intensified human traffic, and an imminent influenza A pandemic the knowledge of dynamical and statistical properties of human dispersal is of fundamental importance and acute [7,9,10]. A quantitative statistical theory for human travel and concomitant reliable forecasts would substantially improve and extend existing prevention strategies. Despite its crucial role, a quantitative assessment of human dispersal remains elusive and the opinion that humans disperse diffusively still prevails in many models [11]. In this chapter I will report on a recently developed technique which permits a solid and quantitative assessment of human dispersal on geographical scales [12]. The key idea is to infer the statistical properties of human travel by analysing the geographic circulation of individual bank notes for which comprehensive datasets are collected at the online bill-tracking website www.wheresgeorge.com. The analysis shows that the distribution of travelling distances decays as a power law, indicating that the movement of bank notes is reminiscent of superdiffusive, scale free random walks known as L`evy flights [13]. Secondly, the probability of remaining in a small, spatially confined region for a time T is dominated by heavy tails which attenuate superdiffusive dispersal. I will show that the dispersal of bank notes can be described on many spatiotemporal scales by a two parameter continuous time random walk (CTRW) model to a surprising accuracy. To this end, I will provide a brief introduction to continuous time random walk theory [14] and will show that human dispersal is an ambivalent, effectively superdiffusive process.
Hub-and-Spoke Network Alliances and Mergers: Price-Location Competition in the Airline Industry
Researcher(s): Nicole Adler, Karen Smilowitz
Year: 2007
This paper presents a framework to analyze global alliances and mergers in the airline industry under competition. The framework can help airlines identify partners and network structures, and help governments predict changes in social welfare before accepting or rejecting proposed mergers or alliances. The research combines profit-maximizing objectives to cost-based network design formulations within a game theoretic framework. The resulting analysis enables merging airlines to choose appropriate international hubs for their integrated network based on their own and their competitors’ costs and revenues in the form of best response functions. The results of an illustrative example suggest that some mergers may be more successful than others and optimal international gateway choices change according to the number of competitors remaining in the market. Furthermore, although the pressure on airlines would suggest a strong preference for mergers or alliances, perhaps surprisingly, the solution outcomes whereby all airlines merge or ally are not equilibria in the overall game.
Methodology for Transportation System Redundancy Analysis in the Greater Baltimore Region
Researcher(s): Hani Mahmassani
Year: 2007
Dr. Mahmassani served as PI of this study, funded by the Baltimore Metropolitan Council. This study developed a regional dynamic network model for simulation-assignment applications to examine the ability of the transportation network and services to withstand shocks and disruptions resulting from natural or man-made hazards and events, ascertain the extent to which the system would be able to meet the mobility needs of the Greater Baltimore Region residents and businesses, and develop contingency measures and strategies to cope with the resulting travel demand patterns under constrained supply conditions. The project provides an example of how to build a large scale simulation platform given existing planning network model.
Solving the Dynamic User Optimal Assignment Problem Considering Queue Spillback
Researcher(s): Yu Nie, H.M. Zhang
Year: 2007
This paper studies the dynamic user optimal (DUO) traffic assignment problem considering simultaneous route and departure time choice. The DUO problem is formulated as a discrete variational inequality (DVI), with an embedded LWR-consistent mesoscopic dynamic network loading (DNL) model to encapsulate traffic dynamics. The presented DNL model is capable of capturing realistic traffic phenomena such a queue spillback. Various VI solution algorithms, particularly those based on feasible directions and a line search, are applied to solve the formulated DUO problem. Two examples are constructed to check equilibrium solutions obtained from numerical algorithms, to compare the performance of the algorithms, and to study the impacts of traffic interacts across multiple links on equilibrium solutions.
REORIENT: Implementing Change in the European Railway Area
Researcher(s): Hani Mahmassani
Year: 2006
Dr. Mahmassani led, with Dr. Elise Miller-Hooks of the University of Maryland, a $1.06Million effort as part of a $7.2M multi-national multi-partner consortium project. The REORIENT project assessed the process of transforming the European railways from nationally fragmented into internationally integrated rail operating systems as a consequence of the EC interoperability legislation. By so doing, it supported the EU policy of balancing modal split between road and rail freight transport.
The team led by Dr. Mahmassani, jointly with Dr. E. Miller-Hooks at the University of Maryland, had lead responsibility for developing and validating strategies for identifying and removing technological, cultural, social and managerial barriers facing the implementation of competitive intermodal rail freight services across national boundaries. As such, the team was in charge of developing the key recommendations that arose from the entire research effort. The recommendations are necessarily based on a comprehensive understanding of the operational, institutional and political context surrounding freight service in Europe. Sophisticated quantitative and qualitative analyses of the operational and social aspects of the freight system likewise compose another essential basis for essential basis for any recommendations.
Dr. Mahmassani’s team also served as coordinator of all network modeling activities needed to support the project, and led the process of building the corresponding network models and associated freight flow processes through the rail network links and intermodal transfer points, as well as the demand models for short and long term freight flow in the study area. This resulted in development of a novel network modeling platform to support evaluation of different strategies and measures intended to improve the prospects of rail freight in the corridor, as well as improvement of capacity and service levels. As such, the network modeling effort plays a critical role in supporting the business case development. In addition, the project led to development of novel ways of calibrating and estimating demand models combining various data sources at both macroscopic and microscopic levels. These activities cut across several other work packages led by other entities. The project also involves coming up with a collaborative decision-making framework by which different entities in different countries, including private service providers, can jointly manage complex systems in real time.
Using Simulation to Test Traffic Incident Management Strategies: The Benefits of Preplanning
Researcher(s): John J. Wirtz, Joseph L. Schofer, David F. Schulz
Year: 2005
This study tested a dynamic traffic assignment model as a tool for pre-planning strategies for managing major freeway incidents. Incidents of various scales and durations were modeled in the northern Chicago, Illinois, highway network, and the impacts of incidents and response actions were measured in lane mile hours of highway links at Level of Service F and spread of congestion to alternate routes around the incident. It was found that the best response action to a given incident scenario was not necessarily intuitive and that implementing the wrong response could worsen congestion on the directly impacted freeway and its surrounding highway network. The simulation model showed that a full closure of the freeway caused congestion to spread to alternate parallel routes around the simulated incident. An event of this scale constitutes a major disruption that may warrant handing off traffic control authority from first responders to a corridor or regional traffic management center. Major arterials accessible from the impacted freeway sometimes need increased capacity to provide access to less congested parallel alternate routes during incidents. The simulation model showed that congestion increases with delayed response, underscoring the benefits of preplanning to speed the implementation of effective incident response actions. Regression analysis using data generated by the simulation demonstrates that incident scale and duration are statistically significant predictors of lane mile hours of congestion in the zone near the incident and on the expressway.