2013-2014 Projects
The project will be informed by lessons derived from a previous multi-year project at the University of Michigan that measured regional accessibility among the top 50 metropolitan regions of the United States, and will proceed in four interrelated stages:
A. Research political and technical barriers to the adoption of accessibility-based evaluation.
B. Develop case studies of accessibility- and mobility-based evaluation.
C. Develop approaches to incorporating accessibility-based evaluation.
D. Collaborate with practitioners and decision-makers on assessment of accessibility-based evaluation.
The utility of approaches described above will be assessed through interviews and focus groups with practitioners.
TRID Database: https://trid.trb.org/View/1447164
Documents:
TRID Database: https://trid.trb.org/View/1428239
Documents:
Rapidly evolving transportation and energy technology is opening up a tremendous number of possibilities for simultaneously achieving environmental sustainability, economic development, and energy security, but the many possibilities for their interaction greatly complicate analysis to understand the best policy options and strategies for individuals and companies to take to maximize opportunities.
This project proposes to explore the effects of various vehicle design options and more detailed vehicle behavior on the integrated transportation and energy system, with particular interest in studying the effect of vehicle design options on traffic system behavior and fuel and electricity use. The study proposes augmenting the use of MPO data to incorporate the detailed behavior of drivers, new vehicle capabilities, and advanced information systems. The resulting agent based model will be suitable for investigating transportation system behavior under next generation systems and its interaction with the energy system.
TRID Database: https://trid.trb.org/View/1482207
Documents:
This research seeks to estimate time-dependent intersection turning proportions using partial link traffic counts and observed turning proportions provided by heterogeneous sensor technologies.
It will help transportation/highway management agencies determine a desirable sensor deployment plan in terms of how to prioritize the critical links for different sensor characteristics under an annual budget constraint. It will also illustrate that interdependencies arise between information and infrastructure in relation to the vehicles, and that they lead to complexities that require solutions as technology is increasingly leveraged in conjunction with the limited budgets. In terms of broader significance, several methodological approaches involving network-level solutions developed to leverage ITS technologies have been previously limited in terms of real-world deployment due to the unavailability of such O-D matrices. Hence, the proposed research has key practical implications for transportation agencies.
TRID Database: https://trid.trb.org/View/1482209
Documents:
Recent advances in real-time traffic sensing, including GPS data from probe vehicles, automatic vehicle identification using RFID and Bluetooth sensors, and automatic number plate recognition, provide richer data when combined with traditional O-D estimation techniques. However, the data obtained from these different sensors do not convey similar information on the traffic conditions of the network. This project seeks to develop and test a systematic methodology to integrate the different data sources, also labeled data fusion, to address the O-D estimation problem, leveraging the availability of different types of data with disparate characteristics.
The study will involve collecting data from ITS test-bed corridors in Chennai (Madras), India. The data collected will also serve as a benchmark data archive for O-D estimation techniques and will augment ongoing research to develop dynamic O-D demand matrices based on partial observability of the field network.
TRID Database: https://trid.trb.org/View/1482208
Documents:
TRID Database: https://trid.trb.org/View/1459032
Documents:
The research is designed to test:
a. The sensitivity of the long-term decision of residential location choice to information;
b. The sensitivity of short-term travel behavior to long-term residential location choice.
Rare among policy investigations, information-related questions can be researched through true experimental designs. This project proposes to assign movers in the two cities randomly to control and experimental groups. Experimental groups in Ann Arbor, MI and Lafayette, IN will be exposed to an information-delivery strategy designed to address transportation-relevant decision-making over a range of time scales. Control and experimental groups will be surveyed for transportation-related outcomes, and intergroup differences will be analyzed with standard statistical models to determine treatment effects.
This project proposes to study how travelers' long- and short-term transportation-related decisions are affected by information interventions, providing potentially new perspectives to fostering sustainable transportation choices and bridging methodological gaps in holistically approaching the notion of livability.
Current strategies for the dissemination of transportation information concentrate at the short-term end of the spectrum. Not enough is known about the sensitivity of choices at varying time scales to information interventions or about the downstream impact of longer-term choices on those made over the shorter term. This project will develop new decision-making models informed by multiple disciplines, including cognitive science, behavioral economics, marketing, transportation, and urban planning. The project will design information interventions intended for the full range of transportation-relevant decisions and test their impacts on people moving to Ann Arbor, Michigan, and West Lafayette, Indiana, as well as consumers in the market for a vehicle.
TRID Database: https://trid.trb.org/View/1447161
Documents:
TRID Database: https://trid.trb.org/View/1490122
Documents:
Vehicle classification data are used in many transportation applications, including: planning, pavement design, environmental impact studies, traffic control, and traffic safety. Every state in the US maintains a network of vehicle classification stations to explicitly sort vehicles into several classes based on observable features, e.g., length, number of axles, axle spacing, etc. Various technologies are used for this automated classification, the three most common approaches are: weigh in motion (WIM); axle-based classification from a combination of loop detectors, piezoelectric sensors or pneumatic sensors; and length-based classification from dual loop detectors. All of these sensor technologies suffer from the difficulty of deploying and maintaining in/on pavement sensors. There has recently been an increasing interest in developing non-intrusive sensors to classify vehicles, e.g., there are several non-intrusive sensors now on the market that offer vehicle classification and most of these sensors rely on microwave radar (e.g., RTMS, SmartSensor, etc.).
The research will deploy LIDAR based system using high vantage points (10-30 m) at one or more multi-lane facilities to monitor traffic and overcome the current limitation due occlusions. In addition to algorithm development, the research will include extensive, labor-intensive ground truth data extraction, both for development and validation of the algorithms. The budget and scope of the work is for the task of developing the LIDAR based system.
Documents:
This work will develop a data-driven mapping platform for assessing local and regional economic development, employment, entrepreneurship, and industry cluster development opportunities related to New Mobility, starting with southeast Michigan as the initial prototype area. The mapping activity will assemble through a GIS-based toolset, a dynamic visualization and geospatialization platform to illuminate existing material and economic flows between related sector agents, as well as to identify network gaps via methodologies related to value-chain mapping.
In general, this suite of network visualization tools relates specifically to New Mobility industry and enterprise, fills an important analysis gap and will result in more informed decision making and innovation by governments, large business, entrepreneurs and other innovators. The work addresses the specific interests of project partners by integrating diverse sets of data and interrelations that operate within ‘blind spots’ of individual sector participants. The tools proposed for development aim to result in more informed decision-making and risk assessment in emerging sectors, while identifying the priority strategic actions Michigan policy and business leadership can advance that support the growth of the new mobility industry sector in Michigan. The proposed tools will be developed with the intent that process and product can be scaled and translated to other sites and globally.
TRID Database: https://trid.trb.org/View/1447163
Documents:
TRID Database: https://trid.trb.org/View/1428240
Documents:
Roadway congestion impacts almost all aspects of our lives in the US, from safety, to the environment, to the quality of life, to the cost of goods and services. A comprehensive understanding of the traffic conditions over space that give rise to congestion remains elusive. To date, these issues have been studied predominantly with macroscopic data from point detectors (e.g., loop detectors) aggregated over fixed time periods ranging from 20 sec to 15 min. Many new theories have emerged in recent years to explain several on-going anomalies in traditional traffic flow theory. At the core of these new theories is the presence of non-trivial disturbances that last far less than the fixed time aggregation periods commonly used to study traffic, and thus, these micro-disturbances have not been empirically observed. If these theories are proven empirically, they should lead to better congestion management and control.
The proposed research seeks to develop the tools to measure traffic flow at a resolution sufficiently precise to measure the micro-disturbances and prove or disprove the traffic flow theories that depend on their presence. Under support from NSF and FTA, OSU has developed an instrumented probe vehicle that includes positioning sensors (DGPS and inertial navigation) and ranging sensors (six LIDAR, one radar). The focus of the RNS is the one forward facing and one rear facing LIDAR sensors. These LIDAR collect a rich, 180° scan out to 80 m, in a plane approximately 0.5 m above the roadway, at 40 Hz. Although hundreds of hours of data have been collected, the tools to automatically reduce this vast quantity of data to useful information still need to be developed. The proposed research would undertake the task of segmenting the vehicle returns from the non-vehicle objects in the LIDAR data, grouping the vehicle returns into discrete vehicles, and tracking the resulting vehicle groups across scans. Once these tools are developed, they would be used to mine hundreds of hours of existing instrumented probe vehicle data.
TRID Database: https://trid.trb.org/View/1490126
Documents:
TRID Database: https://trid.trb.org/View/1447162
Documents:
The project aims to develop a data processing module for a novel LiDAR-based traffic scanner - TScan. The TScan is being developed by the Center for Road Safety to collect microscopic highly accurate traffic data at road intersections. TScan uses Light Detection and Ranging (LiDAR) technology. This technology can detect various types of road users including buses, cars, pedestrians, and bicycles and, unlike video detection, it does not experience the well-known occlusion problem. The system consists of the LiDAR HDL-64E manufactured by the LiDAR Division of Velodyne Acoustics, Inc. installed on a pneumatic 42-foot telescoping mast elevated above the ground and positioned near a studied intersection. The sensor head rotates 900 times per minute, which results in 1.3 million data points per second. Data collected over a period of several hours to several days is stored in high-capacity devices. The system has been designed and all the components of the TScan system have been purchased or manufactured.
The proposed effort covers the first phase of the overall effort. The second phase (not included in this proposal) will immediately follow the first one and it is meant to demonstrate the quality and usefulness of the obtained data for traffic conflict analysis. This phase will build an interface between the already developed data processing module and the existing Surrogate Safety Assessment Model (SSAM), which is freely available public domain software developed by Siemens ITS with FHWA funding. SSAM will convert the microscopic traffic information produced with TScan into meaningful safety-related information such as traffic conflicts and other risky interactions.
The proposed research component focuses on developing a module capable of converting the source data into the microscopic measurements of the motion of identified objects across the field of view in a way to make it useful for more advanced analysis. Although the HDL64E unit in our possession was found useful for autonomous driving, it is still not clear if it can measure the dimensions and motion of objects at a sufficient level of quality for the envisioned applications. This is our primary research objective besides developing the data process to facilitate the required data conversion.
TRID Database: https://trid.trb.org/View/1330765
Documents:
TRID Database: https://trid.trb.org/View/1447166
Documents:
The main purpose of this research is to present an analysis of an alternative strategy for optimizing transportation infrastructure maintenance decision-making. The approach being proposed is based on performance-based resource analysis, which balances competing objectives and perspectives of multiple stakeholder groups while considering the amount of resources available, as described in NCHRP Report 666. Performance-based analysis focuses on the concept of spending efficiencies and performance-based resource allocation. It also encourages the use of performance targets and the ways to develop and maintain them over time. Performance targets have to become a part of the business process that directly links organizational goals and objectives to available resources and results. In performance-based resource analysis, targets are critical when evaluating the effectiveness of investment decisions.
The primary benefit of this research is that it empowers maintenance administrators in state transportation agencies with a new and innovative integrated solution to make decisions and set policies related to transportation infrastructure maintenance. The result of this research provides an alternative way to look at how efficient maintenance spending has been, and gives maintenance administrators a chance to figure out the most efficient way to allocate and distribute maintenance funds.
TRID Database: https://trid.trb.org/View/1428238
Documents:
TRID Database: https://trid.trb.org/View/1490128
Documents:
This project builds upon the progress made by the NEXTRANS investigators in using APC data from transit buses to estimate route-level OD flows considering a variety of new dimensions and identified limitations. Specifically, these efforts relate to the temporal representation of OD flows.
Route-level bus passenger OD flow estimation methods recently developed by the NEXTRANS investigators are presently being used to provide insights on empirical flow patterns for a few transit agencies. The research here is targeted to improve upon these applications for sustained, long-term use. As was done in moving the recently developed approaches toward empirical implementation, methodological formulations must be developed, evaluated, and refined before being put into use. The advanced methods eventually developed would form the basis for long-term benefits to transit agencies and MPOs.
The methods being developed are based on exploiting spatially and temporally extensive boarding and alighting data that are now available from APC technologies in use on many transit properties. As with previously developed methods, the new methods will be inspired by an understanding of bus passenger behavior that is consistent with data and in situ observations collected on OSU’s living Campus Transit Lab and refined according to these data and observations.
The developed methods will lead to a richer representation of OD flow patterns and more accurate estimates of such patterns. In both cases, improved service planning and operations, where OD flow patterns are used as inputs, are expected. Planning applications include, for example, extending, splitting or combining, and designing new routes, and operations applications include short-turning, expressing, and holding. Improved service and operations will eventually result in a more competitive transit mode, with subsequent effects on reduced congestion, improved sustainable use of energy resources, and mitigated environmental impacts stemming from passenger travel.
TRID Database: https://trid.trb.org/View/1473076
Documents: