Current Funded Research

Development of a traffic Density Estimation Algorithm with Simulation and Analysis of Lane-by-Lane Traffic Chracteristics
Investigator: R. Jayakrishnan
Support: Korea Institute of Civil Engineering and Building Technology

Estimate traffic density is of critical importance to understand current traffic conditions. An algorithm for estimating traffic density using data from automobile sensors has a pivotal role in improving the accuracy of road congestion prediction technologies. Traditional traffic density has been generalized at the link level. But the densities of individual lanes on a link vary when a road become congested, meaning that the congestion level has a correlation with density variations among individual lanes.

This research has three objectives:

  • Develop a traffic density estimation algorithm for use with automobile sensor data.
  • Analyze its effectiveness by using a microscopic simulation approach.
  • Analyze lane-by-lane traffic variations according to congestion levels.
  • The developed traffic density estimation algorithm plays a core function for traffic congestion prediction, which enables establishing efficient traffic operation strategies. The simulation provides the test case implementation environment where a research can conduct various experiments in the laboratory.

    Designing a Transit-Feeder System Using Bikesharing and Peer-to-Peer Ridesharing
    Investigator: R. Jayakrishnan and Michael G. McNally
    Support: University of California Transportation Center

    This proposal is on the second phase of the previously funded UCConnect project entitled "Promoting Peer-to-Peer Ridesharing Services as Transit System Feeders". For the second phase of the project, we propose expanding the transportation alternatives available for the LA Metro red line last-mile challenge by including bikesharing in the transit-feeder system. In light of the Los Angeles County's plans to launch bikesharing programs in downtown Los Angeles and Pasadena in the near future. this proposal investigates the potential contribution bikesharing can have on public transport ridership. In addition. we propose to develop a mobile application that uses a combination of Metro red and gold lines. P2P ridesharing. and bikesharing to provide multi-modal itineraries to users.

    Experimental Studies of Traffic Incident Management with Pricing, Private Information, and Diverse Subjects
    Investigator: David Brownstone and Michael McBride
    Support: University of California Transportation Center

    The effective management of traffic incidents and other irregular disruptions on roadways is key to minimizing travel delay and improving the quality of life for urban residents and businesses. We are currently using economic experiments involving human subjects and a networked, realistic driving simulation to study driver behavior in response to information displayed by variable message systems and to dynamic road pricing schemes. Based on our existing results, we propose four new extensions to our study: the addition of more realistic driving mechanics to test driver responses to our treatments under increased cognitive load, the recruitment of subjects outside the UCI student body to confirm the validity of our results with different demographic groups, the implementation of treatments to study the impact of private information messaging systems (e.g. Waze, Google Maps, etc.), and the implementation of treatments to study a novel value-of-time based auction system for toll lane pricing and allocation. Improvements to the driving realism and the Representativeness of our experimental subject pool will strengthen the robustness and validity of our study's results, while the investigation of private information messaging and value-of-time auction scenarios will shed light on their potential for improving transportation management.

    Evaluation of Signalized Intersection Safety Using Centracs System
    Investigator: Xinkai Wu and Will Recker
    Support: University of California Transportation Center

    This research aims to explore the possibility of using sec-by-sec traffic signal data provided by the Centracs system to evaluate intersection safety. Intersection safety has long become a national concern. However, traditional methods, either using historical crash data collected from infrequently happened collisions, or potential conflicts estimated from microscopic simulation which assumes "accident-free", cannot provide accurate and timely evaluation of intersection safety. By contrast, this research estimates potential traffic conflicts using sec-by-sec data extracted from Centracs, therefore has the potential to provide timely and accurate intersection safety information. Centracs has gained more and more popularity in the nation, especially in California, due to its robust capability for improving intersection efficiency. Especially, Centracs collects and archives sec-by-sec traffic and signal data, which can be used to evaluate safety. Using sec-by-sec data, this research proposes an innovative method to estimate potential traffic conflicts and predict red-light violations, which essentially indicate the safety level of intersections. The proposed research will be tested using the data collected from 5 intersections in Anaheim, CA. This research is expected to contribute significantly to the improvement of intersection safety, and build a foundation for future dynamic systems that could alert drivers of emerging and impending hazardous situations.

    An Activity-based Toolbox for Planning Applications with Special Relevance to Transit
    Investigator: Will Recker
    Support: California Department of Transportation and University of California Transportation Center

    The research objective is to produce an activity-based travel demand toolbox that can be used by planning agencies in practical planning applications, particularly those related to transit.

    In this research, we will develop a comprehensive activity-based travel demand forecast model that integrates different variations of discrete choice models, mathematical programming models of activity scheduling and travel choice, fuzzy concepts and machine learning techniques. The research is designed with the main goal of producing an activity- based travel demand toolbox that can be used in practical planning applications. As envisioned, the toolbox will enable users to predict activity patterns and trip chains at both disaggregate and aggregate levels for a study region, analyze public transportation market share and evaluate the impacts of different policies on travel pattern of individuals. Core codes of the toolbox will be in Matlab and Python, and use Visual Basic for the user interface. The codes will be standalone executable files that have minimum software requirements for execution. The toolbox developed in this research will provide planning agencies with a user-friendly procedure to predict activity patterns and trip chains at both disaggregate and aggregate levels for a study region, analyze public transportation market share, and evaluate the impacts of different policies on travel pattern of individuals.

    The methodology to be used here builds extensively on previous work that has successfully extended the original HAPP model to include such considerations as: ridesharing, transit usage, uncertain activity duration and time window constraints, congestion, and calibration/estimation. Such aspects of the methodology have been the subject of a number of papers reported in the literature and are not repeated here. Rather, we summarize that the methodological issues and generalizations associated with the HAPP modeling system have been addressed rather fully, but remain largely confined to the research literature. Our purpose here is to now take this "research modeling system" and turn it into a comprehensive, and "user-friendly," toolbox for application by planning organizations in investigating and evaluating policies affecting the movements of people as they go about performing daily activities.

    Analysis and Synthesis of Electric Vehicle and Charging Data for a Multi-Mode Mobility System
    Investigators: Scott Samuelsen and Tim Brown
    Support: U.S. Department of Transportation and California Department of Transportation/University of California Transportation Center

    This project will explore PEV use and charging patterns in combination with unique vehicle attributes, to explore mixed-mode mobility systems that leverage PEV performance characteristics while minimizing their limitations. Specifically, this project will utilize data available from the Zero Emission Vehicles Network Enabled Transport program in conjunction with the established Spatially and Temporally Resolved Energy and Environment Tool (STREET) to analyze mixed-mode mobility system utilizing PEVs.

    By better understanding plug-in electric vehicle (PEV) use and charging patterns in combination with unique vehicle attributes, this research will explore mixed-mode mobility systems that leverage PEV attributes while minimizing their limitations

    A Simulation-Based Approach to Quantify Congestion and Air Pollution Impacts from Road Incidents
    Investigators: Jean-Daniel Saphores and R. Jayakrishnan
    Support: U.S. Department of Transportation and California Department of Transportation/University of California Transportation Center

    Understanding the impacts of traffic incidents on congestion and air pollution is critical to better target limited public funds to reduce these transportation externalities. Unfortunately, non-recurring congestion related to incidents and its related air pollutant emissions are not known accurately. The purpose of this project explore ways of better quantifying the impacts of freeway incidents on both congestion and the emissions of common air pollutants (CO, CO2, NOx, PM, and VOC) using vehicle microsimulation (TransModeler) combined with EPA's MOVES model. In addition, we will estimate statistical models that will link incident-caused congestion as well as incident- caused air pollutant emissions to traffic and incident characteristics. Although we will analyze portions of Southern California freeways, our methodology will be widely applicable.

    A Unified Framework for Analyzing and Designing Signals for Stationary Arterial Networks
    Investigator: Wenlong Jin
    Support: California Department of Transportation and University of California Transportation Center

    The objective of this research is to develop a unified theoretical and simulation framework for analyzing and designing signals for stationary arterial networks. Existing traffic flow models used in design and analysis of signal control strategies are either too simple to be realistic or too detailed to be efficient. In this research, we propose to apply the link transmission model to analyze and simulate traffic dynamics in a signalized arterial network. In particular we will (1) analytically derive macroscopic fundamental diagrams for stationary traffic patterns with different network topologies, road conditions, driving behaviors, and signal settings; (2) quantify congestion mitigation effects of different signal settings, including cycle lengths, green splits, and offsets, as well as speed limits and road lengths; (3) formulate an optimization problem with the network flow-rate as performance measure to find optimal road, speed limit, and signal control parameters under certain demand levels, and (4) develop a set of simple decision-support tools for arterial network improvement. We will apply our methods for analyzing and designing signals for a corridor with a state highway and side streets. The research matches Priority Theme 4 and could help Caltrans and local transportation agencies to make decisions for improving "main streets".

    Automated Video Incident Detection System
    Investigator: Will Recker
    Support: California Department of Transportation

    Incident management in the Transportation Management Center (TMC) is an active process involving coordination between multiple agencies and field operations teams ranging from emergency services to maintenance. The process, depicted in figure 1 below, involves four major phases:

  • Identification during which the TMC is notified of events that may require active management
  • Verification during which the event is diagnosed for its causes, characteristics, and severity
  • Response during which actions are taken to mitigate the impact of the event on the system including dispatching of field teams, dissemination, and coordination
  • Monitoring during which the current status of the event is continually evaluated and re- diagnosed to determine whether additional response is necessary and/or when field resources have finished tasks and can be deployed to other events
  • Prior research has demonstrated the importance of early verification for reducing the delays associated with capacity-reducing events. The total delay caused by an incident is at least proportional, if not polynomially related, to the time required to verify a given incident. Generally, this is due to the fact that a response to an event cannot be deployed until the problem is diagnosed, leading to significantly greater queuing as the response is delayed. Once an incident is diagnosed through verification, the response process is well organized and more deterministic. As a result, shortening the verification time is a critical component in the efficient management of incidents. Because the above steps are sequential, early verification is ultimately dependent upon early identification.

    As noted in the flowchart, the identification of incidents arrive through a variety of channels including the California Highway Patrol's (CHP) iCAD system, telephone reports from the public, and the Caltrans Closed Circuit Television (CCTV) system. The first two of these channels are particularly effective for reporting of severe incidents, when emergency phone calls are made in response to accidents. Less severe events, such as stalled vehicles or debris in the roadway have less likelihood of being quickly reported by the public. While these situations tend to produce less severe disruptions to traffic flow, they create conditions that are more likely to produce secondary incidents that may have significantly greater severity. At the same time, active monitoring of every stretch of roadway by a TMC operator is time consuming, tedious, and a generally inefficient use of human resources.

    While the automated detection efforts of the past couple of decades have produced mixed results, there is enough promise in these methods to suggest they may aid TMC operators in improving incident identification time. Using automated methods to highlight potential problems for review by TMC staff has the potential to improve response times to a variety of system events, reduce the occurrence of secondary incidents, and generally improve the efficiency of TMC operations.

    In the scope of work that follows, we propose to evaluate the potential of Automated Video Incident Detection (AVID) systems to improve incident identification and response times. Our experience working with TMC operations suggests that TMC operators tend to have developed a level of efficiency with existing processes such that the introduction of new tools and techniques---no matter the technical potential---can actually disrupt existing processes leading to less efficiency overall. We will therefore proceed with a particular eye on how these AVID systems can be directly integrated into existing TMC processes with a minimum of disruption. At the end of the project, we will have deployed a pilot implementation that demonstrates how a selected AVID system can be integrated into TMC processes and will evaluate the impact this system has on the measured delays of particular classes of incidents over time using the Caltrans District 12 TMC Performance evaluation system deployed at the California Traffic Management Laboratories (CTMLabs) at the University of California, Irvine (UCI).

    Bottleneck effects of lane-changing traffic: A macroscopic approach
    Investigator: Wenlong Jin
    Support: U.S. Department of Transportation and California Department of Transportation/University of California Transportation Center

    The objective of this research is to understand stationary and dynamic bottleneck effects of lane-changing traffic from a macroscopic point of view. It is well known that disruptive lane-changes in merging and weaving areas can cause substantial capacity reduction and drop. In this research, we attempt to develop a coherent theory for such critical bottleneck effects with the help of a new fundamental diagram and corresponding kinematic wave models. With NGSIM data, we first establish lane-changing intensity as a function of road geometry, proportion of lane-changing traffic, and other lane-changing behavior characteristics and then derive a fundamental diagram of lane-changing traffic flow, from which we analyze how the number of lanes and proportion of lane-changing traffic would impact capacity reduction in a lane-changing area. This research can improve our understandings of fundamental properties of lane-changing bottlenecks and their impacts on traffic congestion and environments. Insights obtained from the research could help improve lane-management, variable speed limits, ramp metering, and other management and control strategies.

    Bounded acceleration and capacity-drop at merging bottlenecks
    Investigator: Wenlong Jin
    Support: U.S. Department of Transportation and California Department of Transportation/University of California Transportation Center

    The objective of this research is to prove the conjecture that bounded acceleration rates of vehicles can lead to capacity drop inside a merging area. Capacity-drop is one of the most puzzling traffic phenomena occurring near such bottlenecks as lane-drop and merges. While it has been suspected that such a capacity drop is caused by drivers' acceleration behaviors inside various bottleneck areas, there have been no systematic studies on the relationship between drivers' acceleration process and the magnitude of capacity-drop. In this research, we aim to develop, calibrate, and validate a macroscopic model of acceleration behaviors inside a merging bottleneck and quantify their impacts on capacity-drop. From observed vehicles' trajectories, we will calibrate acceleration rates and distances inside such an acceleration zone and calculate the magnitude of capacity drop using the macroscopic acceleration behavior model. The result will be compared with the observed capacity drop from loop detector data. Such a research can improve our understanding of the mechanism and magnitude of capacity drops at freeway bottlenecks. The knowledge can then be employed towards improving ramp metering, variable speed limits, and other control strategies to reduce congestion and vehicle emissions in a road network.

    California Natural Gas Vehicle Incentive Program
    Investigator: Stephen Ritchie
    Support: California Department of Transportation and the California Energy Commission

    This document presents a proposal by the Institute of Transportation Studies at the University of California, Irvine (ITS-Irvine) to develop, administer and support the California Natural Gas Vehicle Incentive Program (NGVIP), funded through the California Energy Commission (CEC) Alternative and Renewable Fuel and Vehicle Technology Program (ARFVTP).

    Assembly Bill 118 (Nunez, Chapter 750, Statutes of 2007) created the ARFVTP. The statute, subsequently amended by AB 109 (Nunez, Chapter 313, Statutes of 2008), authorizes the California Energy Commission to develop and deploy alternative and renewable fuels and advanced transportation technologies to help attain the state's climate change policies.

    The ARFVT Program has an annual budget of approximately $100 million and provides financial support for projects that:

  • Reduce California's use and dependence on petroleum transportation fuels and increase the use of alternative and renewable fuels and advanced vehicle technologies.
  • Produce sustainable alternative and renewable low-carbon fuels in California.
  • Expand alternative fueling infrastructure and fueling stations.
  • Improve the efficiency, performance and market viability of alternative light-, medium-, and heavy-duty vehicle technologies.
  • Retrofit medium- and heavy-duty on-road and non-road vehicle fleets to alternative technologies or fuel use.
  • Expand the alternative fueling infrastructure available to existing fleets, public transit, and transportation corridors. Establish workforce training programs and conduct public outreach on the benefits of alternative transportation fuels and vehicle technologies.
  • To support the goals of the ARFVTP and promote the purchase of clean, alternative fuel vehicles, the Energy Commission has developed a Natural Gas Vehicle Incentive Program (NGVIP) to help lower the incremental cost of these vehicles. Throughout its existence, the NGVIP has been structured to distribute incentives for natural gas vehicle purchases through Original Equipment Manufacturers (OEMs) and their dealers. In this process, OEMs have applied to reserve a number of rebate vouchers to be distributed to end-users who purchase vehicles with qualifying natural gas drivetrains. The CEC now wishes to alter the structure of the program such that end-users are now responsible for applying for rebates directly, with subsequent rebate payments being sent directly to the end-users after purchase.

    ITS-Irvine proposes in this Agreement to administer and support the NGVIP under this new structure, funded through the CEC's ARFVTP. In addition to administering the incentive program, ITS-Irvine will perform research related to the ARFVT mission.

    California Vehicle Inventory and Use Survey
    Investigator: Stephen Ritchie
    Support: California Department of Transportation and EPA Air Resources Board

    This proposal defines a work plan for the California Vehicle Inventory and Use Survey (CA-VIUS), which will focus on freight related vehicles operating in California, with questions designed to obtain annual freight truck activities, operational characteristics, physical characteristics, and the types of commodities carried at the state level. The survey results are needed for the development and validation of the Emissions Factors (EMFAC) model for the California Air Resources Board (CARB), since its assumptions for base and target years are based on VIUS data including vehicle miles traveled (VMT), and commodity and truck fleet characteristics. The survey results are also expected to yield key insights on the inventory and flow of different types of commodities and truck fleets that are critical for statewide freight travel demand modeling as well as forecasting transportation energy demand by the California Department of Transportation (Caltrans) and the California Energy Commission (CEC). Following the discontinuation of the VIUS at the national level by the U.S. Census Bureau in 2002, it has become a challenge to obtain equivalent data for satisfying the aforementioned needs. Thus, this first phase of the CA-VIUS study is proposed to build an appropriate survey framework as well as to implement a pilot survey in California in an effort to provide the necessary update of freight movement data for the State of California.

    The final deliverables of this project will include the following: the proposed survey sample design with associated supporting methodology and data analysis, digitized pilot survey data with completed survey response sheets, statistical analysis of pilot survey results, freight-related factors generated from pilot survey results, and a final report including a comprehensive survey framework, procedure, pilot survey implementation issues, and the corresponding responses to pilot survey implementation issues. While the pilot survey will only cover a selected sub-region of California based on the developed comprehensive survey framework, the pilot survey is expected to provide guidelines and be combined with later surveys applicable to the entire state of California, including the sub-region of the pilot survey.

    The tasks defined for this project are split into three main phases: the first involves survey design and preparation, the second encompasses conducting the pilot survey, while the final phase involves presenting and analyzing survey results. The survey design and preparation phases consist of tasks including literature review, background information collection, decisions for sampling frame, sample size determination, sample selection procedures, survey questionnaire design and survey method selection. To the best of our knowledge, the population of trucks operating in California, regardless of the state in which the trucks are registered, can be obtained both from the California Department of Motor Vehicles (DMV) and the International Registration Plan (IRP), and these sources will be used for sampling framework decisions. Based on initial analysis, a mail-out and mail-back survey design for instate and interstate trucks is likely to be recommended. The design of survey questions will also involve reviewing the 2002 VIUS questionnaire to ensure consistency between them. The tasks for conducting the survey will depend on the anticipated response rate as well as the reliability of the response results. In order to address the possibility of low response rate and to improve survey results, a pilot survey will be prepared and telephone and e-mail services for answering the respondents' questions regarding the survey will be adopted. The tasks associated with the final phase are the analysis of pilot survey results and generation of useful freight-related factors for stakeholders.

    This proposal is organized as follows. The motivation and relevant work are introduced in the subsequent section followed by a discussion of the proposed survey framework and corresponding sub-tasks. The outline of sub-tasks addresses the list of requirements defined by the CARB and other stakeholders. Finally, a list of tasks is proposed along with an estimated timeline for implementing the project.

    CTM-based optimal signal control strategies in urban networks
    Investigator: Wenlong Jin
    Support: California Department of Transportation and University of California Transportation Center

    The objective of this project is to develop optimal signal control strategies in urban networks based on Cell Transmission Model. Traffic in urban networks is getting more and more congested due to the rapid increase in travel demand. Most of the prevailing signal control strategies are developed for uncongested traffic conditions and cannot work properly when traffic gets congested during peak periods. Furthermore, most of them either consider average vehicle arrival rates or model vehicles as queues, and thus, they fail to capture important traffic flow characteristics such as kinematic waves and fundamental speed-density (or flow-density) relations on a road link. To tackle these problems, in this project, we introduce the cell transmission model (CTM) to simulate the evolution patterns of vehicles on a road link Due to the complexity in the time-discrete control signals at signalized intersections, we developed time-continuous junction models )hich can correctly approximate the discrete junction outflows under different traffic conditions, capacity constraints, and signal settings. For CTM with the continuous approximate models, we formulate a nonlinear optimal control problem, in which signal settings (green splits) are control variables, and the network flow-rate in the macroscopic fundamental diagram is the objective function. This project provides a systematical framework to determine optimal signal settings for urban networks. Insights from this project can help engineers and policy makers better understand how the signal settings, route choices, and demand patterns impact the network performance.

    Development of a New Methodology to Characterize Truck Body Types Along California Freeways
    Investigator: Stephen Ritchie
    Support: California Department of Transportation and EPA Air Resources Board

    A very large proportion of freight movement is transported by trucks, and the value and tonnage of goods are expected to grow over time. Trucks have a significant impact on pavement infrastructure, traffic congestion, energy consumption, pollution and "quality of life". To provide a better understanding of the behavior of freight-related truck movements, it is necessary to obtain comprehensive high-resolution truck data. Higher quality truck data can enable more accurate estimates of GHG and other truck emissions which constitute a significant percentage of mobile sources. Higher quality data allows better management of road infrastructure as decision-makers can relate pavement deterioration to more explanatory factors and make more informed decisions for pavement rehabilitation. Higher resolution truck data that provides some insight into the spatial distribution of body types (and hence commodity types) can also make freight forecasting models more accurate.

    Comprehensive truck classification data offers the ability to address the aforementioned concerns. However, truck classification data is available at limited locations such as weigh-in-motion (WIM) facilities. Prior studies have extended some aggregate qualities of the classification data to inductive loop sensors at Vehicle Detector Stations (VDS), which are much more abundant in number but lacking in classification capabilities. There exists no available method to simultaneously provide detailed axle and body type classification of trucks on any existing sensor location, much less to extrapolate detailed classification data from WIM stations to VDS locations.

    Development of a Weigh-in-Motion Testbed (Planning Phase)
    Investigator: Stephen Ritchie
    Support: California Department of Transportation/University of California, Berkeley

    ITS-Irvine researchers have been at the forefront of advanced developments for speed measurement, truck classification, vehicle re-identification and emissions estimation for real-time traffic performance measurement. We have also been actively involved in the analysis of freight data and database design, including the development of the Online California Freight Data Repository, and are currently involved in developing a statewide freight forecasting model for Caltrans. Another current ITS-Irvine research area is the improvement of vehicle miles traveled (VMT) and speed estimates used in EMFAC with weigh-in- motion (WIM) data as one of the main data sources. This proposal to California PATH is to provide research and technical assistance and support for their proposal to Caltrans on Development of a Weigh-in-Motion Testbed (Planning Phase).

    Enhanced Truck Counts for California
    Investigator: Stephen Ritchie
    Support: California Department of Transportation

    A significant proportion of goods movement is transported by trucks, and the value and tonnage of goods are expected to grow over time. Trucks have a significant impact on pavement infrastructure, traffic congestion, pollution and "quality of life". To provide a better understanding of the behavior of freight-related truck movements, it is necessary to obtain comprehensive high-resolution truck data at key truck corridors within the State of California.

    Objectives: The proposed study comprises three main phases. The objective of the first phase enhances the truck body classification models developed by University of California, Irvine (ITS Irvine) in a separate study funded by the California Air Resources Board (ARB), and to develop algorithms to investigate the potential of anonymous truck tracking using the selected technologies. In the second phase, the objective is to set up the technology at selected Weigh-In-Motion (WIM) and Vehicle Detector Station (VDS) sites for advanced truck data collection. In the final phase, the objective is to develop advanced analytics tools to provide detailed reports on the collected truck data, and perform a shakedown of the system and maintenance of installed locations.

    Study Sites: The data to be used for the enhancement of the classification model development will be obtained from multiple sites in California to generate a dataset that contains a significant sample size of key truck configurations. The collect data will be merged with the truck dataset previously collected by ITS Irvine at four sites in California. The investigation into anonymous vehicle tracking using WIM and inductive will involve a separate short-term data collection effort that will involve concurrent data collection between adjacent WIM sites.

    Analysis: The investigators will use a combination of advanced mathematical models and data analysis tools such as artificial neural networks, data clustering, Bayesian and heuristic algorithms to develop the models in this study.

    Anticipated Results: The implementation of the above objectives is expected to yield data that will help to address a significant data gap in truck movement statistics in relation to body configuration, which has implications for truck function and associated industries. This is especially beneficial in providing the freight movement data for improved calibration and validation of the California Statewide Freight Forecasting Model. Without this data, freight modeling efforts can only be calibrated against aggregate truck volumes instead of detailed truck volumes by truck type. Another advantage is that this data could enhance the Vehicle Inventory and Use Survey (VIUS) data for California by providing spatial and temporal characteristics to the vehicle type classifications.

    Evaluating the Travel and Physical Activity Impacts of the Exposition (Expo) Light Rail Line: Leveraging Transit Investments for Community Livability and Regional Sustainability
    Investigators: Doug Houston and Jun Wu
    Support: U.S. Department of Transportation and California Department of Transportation/University of California Transportation Center

    This research will support analysis of data collected in California's first experimental-control, before-and-after evaluation of a major light rail transit (LRT) investment, the Exposition (Expo) line from downtown to the westside of Los Angeles. The region's ambitious LRT construction campaign will support Senate Bill SB375 goals for greater integration of transportation and land-use planning, but we know little about whether and to what degree new LRT is associated with reduced private vehicle travel and increased transit usage. In Fall 2011, we collected geographically detailed 7-day travel data for 285 households along the corridor using daily trip and vehicle odometer logs and supplemental GPS-based location tracking. We will collect comparable "after" data for the same households in Fall 2012 after the Expo line service begins in Spring 2012. The current proposal will support data coding, processing, and analysis and will inform transit planning and community development by (1) assessing the impact of Expo service on nearby private vehicle travel, transit ridership, and physical activity, (2) identifying neighborhood factors which could enhance the potential positive effects of transit proximity on bus ridership and walking, and (3) demonstrating methods for evaluating the sustainability, travel, and community impacts of major transportation projects.

    Experimental Studies for Traffic Incident Management
    Investigators: David Brownstone and Michael McBride
    Support: California Department of Transportation and University of California Center on Economic Competitiveness in Transportation

    The purpose of this project is to use modem experimental methods to identify messaging and pricing schemes best suited towards mitigating delays from unexpected disruptions.

    Traffic incidents and other unexpected disruptions on roadways lead to extensive delays that diminish the quality of life for those that live and/or work in major cities nationwide. The effective management of these incidents is hindered by an incomplete understanding about how drivers respond to information provided by network operators. We propose using economic experiments involving human subjects and a networked, realistic driving simulation to directly study driver behavior in response to information dissemination and pricing schemes designed to manage congestion in traffic networks. Specifically, our study will focus on two mechanisms of management. the use of variable message systems (VMS) and the use of roadway pricing to induce diversion to alternate routes. Our pilot study demonstrates the ability of our platform to elicit reasonable driving behavior from subjects and will guide the implementation and refinement of our full experiment. Messaging scheme improvements for use with extant VMS infrastructure could provide a low-cost tool for general incident management, while insights into messaging/pricing synergies could provide new strategies for the management of tolled facilities.

    Improving Transportation Performance; The Case of Left Turns [REVISED]
    Investigator: Michael G. McNally
    Support: U.S. Department of Transportation and California Department of Transportation/University of California Transportation Center

    Over the past century, the automobile has evolved to dominate transportation not only from a behavioral perspective but from an infrastructure perspective. Thoroughfares that evolved over millennia to serve many users were transformed in decades to the nearly exclusive use of motor vehicles. The reasons for this evolution are well documented; alternatives to this behavioral dominance, numerous in terms of proposals and promise, are nevertheless constrained by the infrastructural dominance. One option that has not been systematically studied but that has the cost advantage of maintaining current infrastructure while addressing associated performance impacts is a significant reduction in allowed arterial left turns. Such a policy will soon become feasible with the rapid adoption of GPS and traveler information systems that can inform drivers of optimal route choice in restricted networks. The proposed research will use a microsimulation approach to investigate a range of left turn restriction and removal options on sample arterial networks, under a range of driver behavior assumptions.

    Infill Dynamics in Rail Transit Corridors; Challenges and Prospects for Integrating Transportation and Land Use Planning
    Investigators: Jae Hong Kim and Doug Houston
    Support: California Department of Transportation and University of California Transportation Center

    Local and regional planning entities are directing substantial employment and housing growth into transit corridors to achieve the sustainability goals of California Senate Bill 375. Despite the substantial focus on transit investment and infill growth, our knowledge base for understanding near-transit infill land use dynamics remains limited. The proposed research will shed light on whether existing plans will be sufficient to encourage favorable land use changes which reorient growth into transit corridors by examining two critical dynamic processes: (1) transit system improvements/expansions and (2) associated land use changes, particularly infill and redevelopment dynamics. More specifically, the project will (a) develop a historical geo-database of the dynamic changes in transit systems and land use over the last two decades (i.e., 1990s and 2000s) in southern California, (b) identify key transit system and policy factors that can shape and re-shape land use patterns in surrounding areas, and (c) analyze infill and redevelopment dynamics associated with transit system improvements/expansions using a parcel-based land use change model. Results will provide insights into the expected nature and magnitude of impacts of proposed rail transit system improvements on infill and growth and will support on-going efforts to more effectively integrate transportation and land use planning.

    Interregional Travel in California: Assessment, Analysis, and Implications for Emissions
    Investigator: Michael McNally
    Support: California Department of Transportation and EPA Air Resources Board

    The first round of Sustainable Communities Strategies (SCS) development is underway in each of California's Metropolitan Planning Organizations (MPOs) to meet the greenhouse gas (GHG) reduction targets adopted by the Air Resources Board in 2010. In developing the SCSs, MPOs use travel demand models to estimate the regional vehicle miles traveled (VMT), which are in turn used to estimate the GHG emissions for the purposes of SB 375. Existing regional travel demand models in California estimate VMT within their regional boundaries using information such as demographics, land use development patterns, transportation infrastructure network, and other related factors. However, these models are limited in their ability to accurately estimate interregional travel. Interregional travel is defined as crossing a regional or MPO boundary which can be measured in terms of vehicle trips or VMT. Recent studies have shown that interregional VMT can range from 6 — 22 percent of the total regional VMT. Estimation of interregional travel may have a significant impact in estimating the GHG emissions. There are numerous factors that contribute to interregional travel such as the lack of jobs-housing balance within the region, tourism, and so on. Therefore, to successfully implement SB 375, it is critical to understand interregional travel.

    In this scope of work, the contractor shall conduct a comprehensive review of existing methodologies or methodologies that are under development of estimating interregional travel. For each methodology, the contractor shall describe in detail the factors used in modeling interregional travel and how the change of each factor affects the estimation. The contractor shall address the weaknesses and advantages of each of the methodologies. In addition, the contractor will propose a recommended methodology for California MPOs to use for estimating interregional travel as part of the modeling efforts in developing their future SCS.

    Knowledge Management Program and Traffic Training Boot Camp
    Investigators: Will Recker, Craig Rindt, and James Marca
    Support: California Department of Transportation

    The UC Irvine Institute of Transportation Studies (ITS) proposes developing a knowledge management program featuring a traffic analysis and technology boot camp. The basic idea of a boot camp is that the participants will very quickly acquire up-to-date knowledge and expertise for application. Throughout their careers, Caltrans staff will often need to refresh and renew their traffic knowledge. At the same time, within its own organization, Caltrans has traffic analysis and technology experts who understand both the theory and Caltrans best practices better than anyone. We are proposing a short-form integrated course that will tap into Caltrans' own organizational expertise to achieve two simultaneous and complementary purposes. First, the participants in the boot camp will review the fundamentals of traffic flow, and then build on those fundamentals both to raise the level of understanding to that required for practical application of the principles involved and to introduce new techniques and technologies. Second, the contents of this pilot course will form an on-going knowledge management function that will capture Caltrans best practices.

    We are proposing a pilot course, in which the course materials will be developed and presented to a small, select group of senior Caltrans employees. The guiding principle of the course development will be to track the contents of the Highway Capacity Manual (HCM). The educational goal is that participants will build a strong understanding of the fundamentals of traffic theory and analysis. Our primary objective will be to provide a hands-on, real-world understanding of the concepts embedded in the HCM methods. In order to translate the HCM into practical knowledge, as well as to fulfill the knowledge management function of the boot camp course, we will organize the lesson plan around case studies culled from actual Caltrans projects.

    The work that we are proposing will occur in two intermingled activities: 1) course development activities that will put together the pilot course and case studies, and lay the groundwork for future replication of the course, and 2) activities related to teaching the course to select Caltrans employees. In the following sections we first present our current thinking on the course content of the traffic boot camp, and then we propose how we will develop the pilot course and lay the groundwork for future camp sessions.

    Moving from Interesting to Implementable Models for Efficient Transportation Systems Management — Breaking Through the Computing Barrier
    Investigators: Amelia Regan and Michael Dillencourt
    Support: U.S. Department of Transportation and California Department of Transportation/University of California Transportation Center

    In this research we propose to extend a decade or more of research in parallel and distributed computing architecture to work on transportation problems falling into the general category of network design, but with time scales that range from real-time to quasi-real time to quarterly or annual planning. We then propose to extend this work to many other operational problems.

    The research objective is to develop a computational platform that will lead to implementable operational systems for real-time network management problems.

    Operational Improvements to the California Vehicle Activity Database (CalVAD)
    Investigator: Will Recker
    Support: California Department of Transportation and EPA Air Resources Board

    The Institute of Transportation Studies at the University of California, Irvine, developed the California Vehicle Activity Database (CaIVAD) for ARB. CaIVAD merges Caltrans' raw vehicle detection system (VDS) volume and occupancy data with Caltrans' weigh in motion (WIM) data. The VDS data is collected approximately every half mile on urban California highways every 30 seconds, and the WIM data weighs and measures every truck at just over 100 detector stations scattered throughout the state on major truck routes. In addition, estimates of hourly arterial volumes are produced by applying hourly scaling factors to the average annual daily traffic (AADT) volumes listed in the California Highway Performance Monitoring System (HPMS) data.

    This proposal is to further enhance and improve the California Vehicle Activity Database (CaIVAD) tool. There are four main objectives to the proposed work. The first is to develop a quality assurance and quality control (QA/QC) methodology that will let CaIVAD users know how much confidence they should place in the numbers they are using. The second task is to improve the outward-facing CaIVAD website. The third task is to add additional years of data and to deploy a method for continuously adding new data to CaIVAD as it becomes available. The third task will also lay the groundwork for integrating the results of the truck classification project work that is being developed in parallel at UCI ITS, as well as any other vehicle activity projects ARB may undertake in the future. The fourth goal of this project is to investigate the use of cloud computing type resources. CaIVAD contains a lot of raw data, and even with its current three-year extent, it is hitting the limits of what is possible to store on a single server. Moving to the "cloud" or to ARB-hosted cloud-like services is explored in this final task, so as to enable continued growth of CalVAD.

    By the end of this CaIVAD project extension, we expect that CalVAD will successfully migrate from research project into a deployed and useful tool for ARB and other public agencies.

    Performance Analysis and Control Design for On-ramp Metering of Active Merging Bottlenecks
    Investigator: Wenlong Jin
    Support: California Department of Transportation and University of California Center on Economic Competitiveness in Transportation

    The objective of this research is to analyze the performance and design the control parameters for both pre-timed and traffic-responsive on-ramp metering of congested merging bottlenecks. In this research we will (1) quantify the congestion mitigation effects of different ramp metering algorithms at an active merging bottleneck, (2) design control parameters for efficient and robust traffic responsive ramp metering algorithms, (3) identify demand patterns when ramp metering algorithms are effective, and (4) develop a set of simple decision-support tools for ramp metering with both kinematic wave models and microscopic simulations. The research will help Caltrans to make decisions on the necessity, priority, algorithm, and parameter tuning related to ramp metering.

    In this research we aim to address these fundamental issues related to ramp metering systems. We will systematically analyze the performance and design the control parameters for pre-timed and traffic-responsive on-ramp metering of congested merging bottlenecks. Such a study is enabled by the PI and collaborators' systematic and fundamental research into the understanding of the complex interplay among merging, lane-changing, and accelerating behaviors at a merging area: in a series of papers, we have developed kinematic wave models for merging, lane-changing, and capacity drop at merging bottlenecks (Jin, 2010a,b,c; 2013; Gan and Jin, 2013; Jin et al., 2013); these models have been calibrated and validated with NGSDM and PeMS data. hi addition, we have studied the performance analysis and control design problem for lane-drop bottlenecks regulated by variable speed limits (Jin and Jin, 2013; Jin and Jin, 2014), where we have successfully addressed the capacity drop issue at such bottlenecks; methods and insights from these studies can be directly useful for attacking the ramp metering problem, which is more complicated due to the extra merging processes at merging bottlenecks.

    In this research we will (1) quantify the congestion mitigation effects of different ramp metering algorithms at an active merging bottleneck, (2) design control parameters for efficient and robust traffic responsive ramp metering algorithms, (3) identify demand patterns when ramp metering algorithms are effective, and (4) develop a set of simple decision-support tools for ramp metering with both kinematic wave models and microscopic simulations.

    In addition to potential theoretical contributions, this research will lead to a set of decision- support tools that can help to answer a series of questions: Is a ramp meter warranted at a location? Which merging areas should be given higher priorities to install ramp meters given a limited budget? Should pre-timed or traffic responsive metering algorithms be implemented? What kind of control parameters lead to more efficient and robust control?

    Promoting Peer-to-Peer Ridesharing Services as Transit System Feeders
    Investigator: R. Jayakrishnan
    Support: California Department of Transportation and University of California Center on Economic Competitiveness in Transportation

    Peer-to-peer ridesharing services are a recently emerging travel option that can help accommodate the growth in urban travel demand, and alleviate some of the current problems such as excessive vehicular emissions. Prior ridesharing projects suggest that the demand for ridesharing is usually shifted from transit, while its true benefits are obtained only if the demand shifts from private autos. This project studies the potential of efficient real-time ride-matching algorithms to augment demand for transit by reducing private auto use. The Los Angeles Metro red line is considered for the case study, since it has recently shown declining ridership. A mobile application with an innovative ride-matching algorithm will be developed as a decision support tool that suggests routes that combine ridesharing and transit The app also facilitates peer-to-peer communications of users via smartphones. For successful ride-sharing, strategically selecting transit stations is crucial, along with the pricing structure for rides. These can be adjusted dynamically based on the feedback from the app-users. A parametric study of the application of real-time. ride- matching algorithms using simulated demand in conjunction with the SCAG model for the selected study area is proposed, along with a limited field study of the peer-to-peer use of the apps.

    This research is designed as the first phase of a longer-term research effort to promote ridesharing as a demand generator for transit, and as such the main results from this first-year research will be definitive plans and applications for elaborate field trials of the possibility. A key deliverable in this project is a real-time multi-hop peer-to-peer ridesharing mobile application. The application is to match drivers (including the transit system) and riders in a ridesharing system with strategically located stations. The locations of the stations in the app would be customized to be compatible with the section of the transit system identified in task 1, but the general platform of the app would be flexible enough to be used with any transit system (or none) once the stations are identified. For the app purposes, we will further develop a ride-matching algorithm that is computationally efficient. The research will provide insights from a simulated study of station efficiency and potential demand-augmentation from multi-modal ride-sharing opportunities with transit systems, as well as an analysis of the sensitivity of the results to pricing. Insights from the preliminary field test will also be provided in the final report, along with a detailed plan for an elaborate subsequent field study. In addition, several academic papers are expected to be submitted to well-known transportation journals.

    Proposal for Advancing the Value of the CHTS to Caltrans
    Investigators: Jean-Daniel Saphores, Craig Rindt, and Suman Mitra
    Support: California Department of Transportation

    University of California, Irvine, Institute of Transportation Studies proposes to provide support to Caltrans to enhance the value of the 2010-12 California Household Travel Survey (CHTS). The 2010-12 CHTS, which resulted from a statewide, collaborative effort, enabled the collection of travel information from 42,560 Californian households. This rich dataset has helped update regional and statewide travel models, but it could also inform Caltrans planning efforts. As such it should be of interest not only to various state and transportation planning agencies across California, but also to staff from the California Department of Transportation (Caltrans). However, the potential value of the CHTS is not always well understood by Caltrans staff. Moreover, some Caltrans staff from the Office of Travel Forecasting and Analysis may benefit from updating their knowledge of statistical modeling to comfortably query CHTS data and to estimate some common transportation econometrics models.

    In this context, we are proposing to: 1) perform a systematic diagnostic review of the 2010¬12 CHTS database for unlikely observations; 2) interview headquarters and district Caltrans staff in three (3) selected Caltrans Districts to better understand how they could benefit from using 2010-12 CHTS data and to help promote the use in their work of CHTS data; 3) provide hands-on statistical training and consulting to selected Caltrans staff in the Office of Travel Forecasting and Analysis in Sacramento and possibly to some district Caltrans staff (for a maximum of twelve (12) staff); 4) provide on-call statistical support to Caltrans staff from the Office of Travel Forecasting and Analysis; and 5) create a reference book of useful statistical commands based on actual case studies to make it easier to put the 2010-12 CHTS to work for Caltrans staff.

    The work we are proposing will start during 2014 with visits of three (3) district offices to explore how CHTS data could be promoted to planning and modeling staff in Caltrans districts. Once there is a clear understanding of District and HQ staff needs, training material will be developed to deliver training modules to staff in the Office of Travel Forecasting and Analysis. The content of the training modules will be determined according to the findings from Headquarters (HQ) and District office visits. Training will be delivered at UC Irvine and in Sacramento. Finally, over the course of this project, a reference book of statistical techniques with Caltrans-based examples will be compiled.

    Rail and the California Economy
    Investigator: David Brownstone
    Support: California Department of Transportation and University of California Center on Economic Competitiveness in Transportation

    Task 1: This phase will begin with a kick-off meeting between Caltrans, the research team, and CalSTA. During this phase, there will be interactions between Caltrans and the research team to advise on the key individuals to consult. Based on the information gathered in the exploratory phase, the research team will work with Caltrans to refine the approach for the remainder of the study. This includes review of the ten areas of impact discussed above in Section 1, which may be adjusted within the limits of the study budget.

    Task 2 : Corridor Case Studies:

  • Market analysis to identify the predominant passenger and freight movements in the corridor for which rail service is used;
  • The impact of rail connections to the major port terminals of Long Beach/Los Angeles and Oakland and potential application to smaller ports lacking rail connections such as Redwood City.
  • An examination of the costs and benefits of the subsidized Amtrak Thruway bus service which connects smaller rural areas with Amtrak California trunk lines.
  • Task 3: Statewide Analysis:

  • Rail market shares and traffic volumes for various categories of passenger and freight movements;
  • Employment and income of rail-dependent businesses.
  • Task 4. Future Impact of Rail on the California Economy:

  • Estimate or determine from existing literature elasticity's required to assess impacts of population and income growth in rail corridors, changes in the State's industrial mix, impacts of the aging of the population, and changes in fuel prices.
  • The Effectiveness of State and Local Incentives on Household Ownership of Alternative Fuel Vehicles - A SEM Analysis
    Investigator: Jean-Daniel Saphores
    Support: California Department of Transportation and University of California Center on Economic Competitiveness in Transportation

    The purpose of this project is to analyze the impact of state and local incentives on household ownership of alternative fuel vehicles (AFVs) and hybrid electric vehicles (HEVs) using generalized path analysis and structural equation models (SEM) while accounting for residential self-selection and demographic characteristics. Incentives include parking privileges, HOV access exemption, income tax credit, sales tax, and rebates. Two public datasets will be analyzed: the 2009 NHTS and the 2012 CHTS. They will be supplemented with household location information already available to the PI (via request to FHWA for the NHTS and thanks to a Caltrans contract for the CHTS), land use data, and information about incentive programs. To correctly account for incentive impacts, we will restrict our analysis to households who purchased a vehicle during the year when either survey was conducted. Understanding the effectiveness of various government policies is important at a time when there is increased interest in promoting AFVs/HEVs to address our dependence on foreign oil, air pollution and global warming.

    Transit, Traffic, and Affordable Housing: Challenges and Prospects for Aligning Housing Policy with Compact Development Goals in Transit Corridors
    Investigators: Doug Houston and Victoria Basolo
    Support: U.S. Department of Transportation and California Department of Transportation/University of California Transportation Center

    The alignment of transportation, land use and housing policy has gained increased importance in California with requirements in Senate Bill 375 for transit-oriented, mixed-use and infill development which accommodates the housing needs of all economic groups. These compact communities could reduce auto dependency, promote more active travel and transit use for low-income groups, but they could also result in greater exposure to traffic and associated air pollution. Several place-based affordable housing programs, such as LIHTC and HOPE VI, encourage development in transit corridors but we know little about whether units built under these programs and units proposed in city housing elements per state law are located in areas with compact development amenities and/or potential traffic-related hazards. The proposed research will evaluate the distribution of existing and proposed place-based units in terms of transit accessibility, nearby land use mix, neighborhood walkability, and traffic exposure, and will compare these distributions with units subsidized through the Housing Choice Voucher program, which is a non-place-based strategy to allow recipients to rent in the private market. Finally, we will conduct case studies to illustrate the challenges, prospects, and distributional implications of using existing affordable housing policies to meet compact development goals in transit corridors.

    Transportation Emergency Management Framework: Earthquakes and Inter-dependent Systems. Phase II: Application Study of an Emergency Response Framework
    Investigator: R. Jayakrishnan
    Support: U.S. Department of Transportation and California Department of Transportation/University of California Transportation Center

    The full proposal that was initially selected for funding in early 2011 is attached. As directed by UCTC, the first half of the work was later proposed as a Phase-I of the project using partial funding in Jan-June 2012, and the remaining work is being proposed here as the Phase-II of the project for funding from the newly approved UCTC consortium. The period of work in is for 12 months (July 1, 2012 - June 30, 2013). The phase-II research proposed here focuses on the second set of tasks from the attached full proposal (pages 3-12). The main focus areas are:
    • Developing a modeling and analysis framework that can provide input for response agency decisions on managing the transportation system under link failure emergency. The framework is based on a mesoscopic simulation platform of evacuation modeling (ONE-ITS, U.California/ U.Toronto), and routing algorithms will be developed for the type of disruptions under earthquakes.
    • Demonstrating the use of the framework in scenario analysis by selecting earthquakes of various intensities for the Los Angeles area network, and modeling the impacts under a variety of disruption conditions. Modeled management actions will include traffic routing strategies.

    Truck Tour-Based Model for Spatial Disaggregation of California Freight Demand
    Investigator: Stephen Ritchie
    Support: U.S. Department of Transportation and California Department of Transportation/University of California Transportation Center

    Freight transportation encompasses the movement of a wide variety of commodities as well as commercial vehicles on the freight infrastructure, linked to socioeconomic condition and polices. Throughout the history of transportation research, the concept that freight movement is responsible for a large share of the diverse problems in transportation has been accepted, but widespread concerns about environmental impacts such as air pollution, noise, and safety have led to a renaissance of new freight related research. In the same vein, the role of statewide freighting forecast models has been expanded into much finer levels of analysis than the county level. In partnership with state agencies and Metropolitan Planning Organizations (MPOs), the California Department of Transportation (Caltrans) is currently developing a California Statewide Freight Forecasting Model (CSFFM). A critical challenge is to provide a framework for organic integration between the CSFFM and a finer spatial level of models to meet MPO needs. However, factoring methods are currently largely used for disaggregating freight demand. Such methods cannot adequately capture the complex structure and behavior of freight movements, advances in logistics, information technology, and relocating infrastructure at the MPO level. One advantage of the CSFFM, modal path-based OD representation, cannot be fully utilized by MPOs because factoring methods tend to break the chains of modal path-based information in the conversion to trip-based information. This research will explore and develop truck tour-based models for disaggregating California Statewide Freight Demand from an aggregate Freight Analysis Zone (FAZ) level to the more disaggregate Traffic Analysis Zone (TAZ) level, by using truck GPS data. Expected results include new and improved insights into the spatial and temporal operations of trucks at the urban and MPO level, contribution to the statewide-related component of urban freight modeling, and an evaluation of traffic and environmental impacts of state-level policies and air pollution mitigation strategies.