Research Areas


Designing a transportation data-warehouse prototype for the Buffalo-Niagara region and demonstrating its usefulness through a specific application.
Analyzing traffic violations and traffic crash records to develop a probabilistic model that will help detect high risk drivers with the main goal of preventing future crashes.
Examining in-vehicle and infrastructure-based technologies to assess how they might impact emergency responders, particularly EMS.
Exploring historical incident and traffic data to revolutionize response strategies.
Creating a mobile computer application for documenting and sharing data regarding vehicular accidents.
Exploring the potential for using a number of machine learning and data mining methods to analyze accident data.
The project investigates how real-time conditions interact to affect driver safety performance changes. From that understanding, practitioners and drivers can make more informed decisions to reduce the likelihood of a crash.
Recently, there has been an unprecedented interest in Connected and Automated Vehicles (CAVs) or self-driving vehicles. CAVs have the potential to revolutionize transportation, resulting in a major paradigm shift in the way we move and move our goods. The current project is conducted in synergy with another project at UB, funded by New York State Energy and Research Development Authority (NYSERDA) and New York State Department of Transportation (NYSDOT). That project is evaluating the technical feasibility, safety and reliability of using CAV technology, and in particular the self-driving shuttle, Olli, manufactured by Local Motors. 
Conventional travel demand and other planning data sources provided very limited coverage on non-motorized modes such as biking and pedestrian. Crowd-sourcing approach has the potential to collect more up-to-date data for these modes with minimal costs and at a continuous basis. However, such data is mostly self-reported and lacks a unified format and standard, which compromises the data quality. More advanced data processing, cleansing, and integration methods are needed to make such data sources useful and valuable. This study investigated a set of biking incidents data collected in the Washington D.C. metropolitan area to explore such potentials.