Project Objective

The main product from the project is a tool by which NYSDOT and NYCDOT can quantify NRD, for specific locations and corridors and for the City in total. The tool has to predict NRD in a way that tracks to the causal factors: for example, the type of incident, location, weather conditions, v/c (volume-to-capacity) ratio, LOS (level of service), vehicle speeds, number of lanes and ramps involved, etc.

Project Abstract

The project will help NYSDOT and NYCDOT better quantify and predict the non-recurring delay (NRD) from incidents on the City?s highway network. NRD is a substantial portion of the total delay that occurs within the City and it can be mitigated through various actions. If NYSDOT and NYCDOT can better predict NRD and trace it to the causing factors, they can take actions to reduce it. Those actions range from TSM and ITS measures to geometric changes and capacity investments.

Task Descriptions:

Task 1: Goals and Objectives Goals, objectives, milestones, and measures of effectiveness (MOE?s) are needed to conduct the NRD analysis. The purpose of this task is to identify what they should be. The goals should be statements of intent (e.g., reduce NRD); the objectives and milestone values should relate to these goals (e.g., achieve 10% reduction in NRD in 2 years); and the MOE?s allow a way to unambiguously assessing those attainments. Candidate MOE?s include vehicle hours of delay (VHD), person-hours of delay (PHD), and facility-hours of delay (FHD). The MOE?s have to be derivable from the data collected for purposes of adopting / adapting the tool to NYC conditions.

Task 2: Inventory NRD is not a new phenomenon. Other urban areas have been trying to measure and predict it. In this task, these existing methodologies and tools will be reviewed. Items of interest will be the methodology employed, the input parameters used, and the strengths and weaknesses of the resulting predictive model.

Task 3: Data Collection and Data Analysis Once the goals, objectives, milestones, and MOE?s have been set (Task 1) and the set of possible models has been identified, data need to be collected and analyzed so that the best possible NRD model can be developed for NYC conditions. This is a three-step process, divided between two tasks. The steps are a) data assembly, b) data analysis/model development, and c) coefficient estimation for the look-up tables. The RFP places the first and second of these steps in Task 3 while the third is in Task 4. That delineation has been kept here.

Task 4: Development of Look-Up Tables The main goal of the overall project is to develop a set of Look-Up Tables. They will allow NRD to be estimated based on outputs from the CNAM and possibly other models. The purpose of this task is to develop those look-up tables. The models on which the tables are based will be specified in Task 2. The MOE?s and related goals, objectives, and milestones are from Task 1. The input data for developing the look-up tables come from Task 3. As suggested previously, it might be useful to overlap Tasks 3 and 4 rather than do them in series so that the tables take maximum advantage of the data collected and so the data collection is more sharply focused.

Task 5: Strategy Assessment A variety of models could be linked to the look-up tables. One candidate is the CNAM. Each model has strengths and weaknesses regarding its ability to interface with the look-up tables and predict NRD. The purpose of this task will be to review these alternatives (e.g., CNAM, Tranplan, Paramics, and TransCAD) and evaluating their ability to support the NRD prediction task. A list of models to review will be developed through a dialog with NYSDOT and NYCDOT (and potentially other agencies, such as NYMTC and PANYNJ). The method for assessing the model?s performance and the criteria to employ will also be jointly identified. Access to the models (perhaps copies of them) will be provided by NYSDOT and or NYCDOT in the event that the research team does not have a copy so that the review can be conducted. For purposes of assessment, two time periods will be employed: present conditions and those anticipated for six years in the future. The resulting deliverable will be a TM summarizing the results of the assessment.

The modeling tools developed in this investigation will be used to quantity the NRD that now occurs and predict what it will be six years into the future. This modeling exercise will also consider the estimation of NRD for both the base case condition and the different strategies considered.

Task 6: Report Preparation A final report is needed that describes the results of the project. The report will include the tables which when used in conjunction with the approved (selected) simulation models will predict local and system-level NRD. The report will be delivered in WordPerfect / QuatroPro format. The medium will be 5 CD-ROM copies and 25 printed copies.

Budget

$317,751

Student Involvement

This project will involve several graduate and Doctoral student in data collection, presentations, literature documentations and report preparation.

Relationship with Other Research Activities

None

Technology Transfer Activities

See below

Benefits of the Project

The project deliverables will be a series of technical memoranda (TM), one for each task:

Task 1: A description of the NRD-related goals, objectives, milestones and MOE?s selected and the implications these have for the subsequent task efforts

Task 2: A summary of the models and mitigation strategies identified. The summary will use matrices / tables to compare the different models via the MOE?s from Task 1.

Task 3: A summary of the data collected including sources. The data will be organized in a format suitable for quantifying NRD. The impacts of non-recurring incidents will be related to causal factors, possibly using a series of matrices.

Task 4: A set of look-up tables for NRD specifically for NYC that can be used with delay models such as the CNAM. These geographic-specific tables will enable such models to predict the impact of non-recurring incidents (singly and in clusters) in terms of NRD and possibly other metrics.

Task 5: Results from the analysis of available models and methodologies in terms of their ability to interface with the look-up tables and predict NRD.

Task 6: A final report, in WordPerfect / Quattro Pro format: 5 CD-ROM copies and 25 printed documents.

Key Words

Traffic congestion, traffic management, traffic flow, transport modeling