Arterial Performance Measurement Using Multiple Traffic Data Sources

UB researchers are looking into how using multiple data sources can accurately depict real-time performance metrics for arterial roadways.

This research will investigate how the multiple traffic data sources can be integrated in a consistent manner, and how they may be best used for arterial performance measurement. Critical issues associated with this process such as privacy protection will also be studied.

As technologies advance, emerging urban data are increasingly available for wide urban areas. They include data from dedicated loop detectors, license plate readers, vehicle re-identification, electric toll collection (ETC), Bluetooth, cellular phones, GPS-enabled devices (such as navigation devices and smart phones), potentially Connected Vehicles, and so on. Such data are inherently heterogeneous, including both fixed-location data (e.g.., those from loops) and   mobile data (e.g., those from GPS), which are referred to as Urban Hybrid Traffic Data (U-HTD) in this proposal. U-HTD provides great opportunities for urban transportation/traffic system performance evaluation, modeling, and management. They can directly measure how the urban system performs and help pinpoint problem areas (such as bottlenecks). Integrated with other data from surveys / interviews / experiments, they can help better understand how the system components behave and interact with each other. They can also help make more informed decisions about how to allocate resources and investment to manage urban systems more effectively and efficiently. On the other hand, however, U-HTD poses great challenges in data collection, processing, storage, and use.

This research aims to tackle some of these challenges by developing methods on how to best mine the different data elements in U-HTD, how to protect privacy when processing and using U-HTD, and how to developing novel methods that can best utilize U-HTD for critical urban transportation applications. The key research areas are listed as follows:

·         Develop big data analytics methods to mine U-HTD for urban traffic modeling applications

·         Privacy implications when processing/using U-HTD for transportation applications and techniques that may alleviate privacy concerns

·         Methods to understand travelers’ behavior, decision making process and the relationship with built environment

·         Methods to evaluate urban freight performances (such as delivery times at stops, number of stops per tour, etc.)

·         Methods to understand the behaviors of key urban freight system components, their interactions, and the relationship between transportation and land use

·         Freight tour synthesis models that use analytical approaches to estimate aggregated freight demand using secondary data sources such as U-HTD

·         Network optimization models for tour-based, multi-class (especially passengers and trucks) origin-destination demand estimation using secondary data sources such as U-HTD

The team will also make its best effort to collaborate with researchers / centers in other TransInfo partners.  Transportation informatics is a relatively new and emerging area, and excellent research has been conducted at UB and other partnering universities. The team plans to share data, research ideas, and collaborate on research projects with them in the near future.

The team will collaborate with other partners of TransInfo as well as the CITE and Volvo VERF-SUSF centers at RPI to educate and train the next generation transportation engineers and leaders. This includes integrating research results into courses, and recruiting graduate and undergraduate students into cutting-edge research. The team will collaborate with the Hudson Valley Community College (HVCC) and Navajo Technical College (NTC) to improve their programs. HVCC is a two-year institution that mainly serves New York State Capital Region, with 10% Black and 5% Hispanic in its student body. NTC is a technical college serving the Navajo Nation (Native Americans). Specifically, RPI will help HVCC and NTC develop curriculum on Geographic Information Technology (GIT), which covers basic GPS and GIS data processing for transportation applications. RPI also plan to recruit HVCC and NTC students to participate in the proposed research projects through summer internships, designed to expose them to cutting edge research and to motivate them to pursue higher education in STEM. The team will also collaborate with RPI’s EA program to conduct outreach activities to high school students in the Albany New York area.

The team will work closely with federal, state, and local transportation agencies and will make their best effort to convey the findings from this research widely to the transportation agencies, as well as to the industry and the academic community. The team has broad access to the transportation engineering community in New York State and in other states of the country. Over the last several years, team members gave presentations on transportation research to decision makers, engineers, and planners at California Department of Transportation, NYSDOT, Capital Region Transportation Commission (the regional MPO), and other state and local agencies. The team plans to continue such efforts by distributing research findings to agencies through committee meetings, seminars, and presentations. The team also has collaborated with leading industry partners in traffic data integrating and mobile computing, and will leverage this to disseminate research results to the industry. The scientific findings of this investigation will also be disseminated via journal publication, web pages, and professional conferences to research communities in transportation, machine learning, and computer sciences.

Researchers: RPI professors Jeff Ban, Jose Holguin-Veras, and Cara Wang

Partners: NYSDOT, New York City DOT (NYCDOT), and the private sector

Data Sources: (1) DOT traffic monitoring and data collection systems (loops, video cameras, E-ZPass readers, etc.); (2) mobile sensing data from the private sector; (3) blue-tooth devices deployed by the research team.