What is Transportation Informatics?

With nearly a billion automobiles on the road today, and a doubling projected over the next decade, transportation provides indispensable functions in support of meeting the basic needs of society for mobility and accessibility, and directly affects the quality of life. However, with the continued increase in travel demand and non-sustainable development patterns, transportation systems have begun to show signs of serious strain, such as congestion, traffic accident, excessive energy consumption and increased emission level. To mitigate these problems, aside from developing more efficient and intelligent automobiles, making the acquisition and spread of road/traffic information more precisely, efficiently, and timely is of equal, if not more, importance. 

Highway Safety has been identified as the top priority in US and all around the world. Traffic crashes have been identified as the fourth cause of deaths in the world according to the Center for Disease Control and Prevention (2013) behind hearth diseased, cancer and respiratory problems. In fact, the effects of traffic crashes cost billions of dollars per year in the US alone.

As an extension of the work that was completed for year one funding related to “Developing Highway Safety Performance Metrics in an Advanced Connected Vehicle Environment Utilizing Near-Crash Events from the SHRP 2 Naturalistic Driving Study” CUBRC intends to investigate the application of knowledge discovery (KD) techniques to analyze the data that was compiled in year one in relation to other ‘big’ transportation data sets. The title for the new work will be “An Evaluation of Knowledge Discovery (KDD) Techniques for ‘Big’ Transportation Data.” It is envisioned that CUBRC will complete work under the following tasks.

MAP-21 requires that greater emphasis be placed on developing performance measures to justify roadway safety improvements. This project proposes to develop an alternative highway safety performance measure through the monitoring and classification of near crash events.

Public transit planning is a complex design problem that involves a wide range of research topics and methodologies. The critical input of these problems is the Origin-Destination (OD) demand and the quality of OD demand information is crucial for these complex design and planning problems. The main deficiency of OD demand data format currently used in various transit planning models is that the ODs are defined at transit stop-level. By using stop-level ODs, a decision-maker neglects the multi-modal nature of transit and loses the information of transit travel demand elasticity and flexibility, which results in suboptimal solutions for transit planning problems. 

This project aims at improving incident response strategies by exploring historical incident and traffic data. Traffic incidents have become a major cause of congestion and significant threat to urban mobility. Many road networks in major cities are currently operating near, if not beyond, capacity during peak hours. Capacity reduction and road closure due to incidents can cause significant delays over an extended period. An effective incident management system not only helps to mitigate congestion through swift incident detection, response, and site clearance, but also generates significant environmental benefits by reducing fuel consumption, emissions, and potential secondary incidents.

The gap between increasing demand of infrastructure and shrinking funds poses great challenges to the governments of different levels. These governments are turning more to the private sector to help make projects both large and small to be built.  It is suggested that private-public partnership (P3) projects can run more efficiently, be finished under budget and ahead of schedule and can have long term goals of maintaining and operating roads, tolls and bridges. However, many arguments in favor of P3 are anecdotal or based on limited number of case studies. Solid empirical studies and quantitative analysis is lacking in literature, most because of the lack of data. The analysis is further confounded by the diversity of projects in scale, time of completion, functional types, geographic distribution, financial sources, contract types, etc.

Travel behavior data enables the understanding of when, where and how people travel, and plays a critical role in travel trend monitoring, transportation planning, and policy decision support. Conventional travel survey provides detailed information on trips in a given time period taken by a representative sample of households within a geographic area. Such survey data has been the primary source of travel behavior information for transportation agencies for a long time, and has played an essential role in the development and maintenance of regional transportation models. The relatively high cost of traditional travel surveys often prohibits frequent survey cycles. Researchers and practitioners desire cost-effective alternatives to the conventional survey method that could provide more up-to-date, or more ideally, longitudinal travel data. There have been some pilot studies that focused on collecting travel data using smartphone apps. However, the potential of smartphone apps in travel data collection cannot be fully unlocked until researchers overcome several major barriers, including challenges in subject recruitment, sampling bias, and privacy concerns.

Social media has become ubiquitous in our daily life and there has been a growing interest in mining useful information therein. Vast of user-generated content strengthens the linkage and interaction between each individual within the community circle, and also provides large amount of information related to various areas. Examples include Facebook, Twitter, YouTube, Google Plus and Wikipedia. The trend of easily accessing social media will continuously grow with the development and commercialization of wearable computer devices, like Google Glass and other smart watches.

Using big data informatics methods, the intent of this study is to first conduct a detailed, multivariate statistical assessment of pavement treatments by PPP type, by studying their performance in terms of extending pavement lives. Also, the elapsed time until the pavement crosses a threshold that is considered critical will be investigated, using hazard-based duration statistical modeling. The findings aim to improve the process by which pavement rehabilitation treatments are planned and implemented at both the project and network levels. To that end, an excel-based application that predicts the pavement’s performance over time, and forecasts the pavement life by treatment type and PPP type, will be developed. The application will be designed to facilitate decision making under big data conditions, in terms of evaluating pavement-treatments’ effectiveness by PPP type, in order to allow the selection of PPP pavement rehabilitation treatment options that will last the longest given funding limitations.

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.

In recent years, and as a result of the continued increase in travel demand across the border coupled with the need for tighter security and inspection procedures after September 11, border crossing delay has become a critical problem with tremendous economic and social costs. 

With the recent advances in data collection, storage and archival methods, the size of accident datasets has grown significantly. This in turn has motivated research on applying data mining and complex network analysis algorithms, which are specifically designed to handle datasets with large dimensions, to traffic accident analysis. This project is exploring the potential for using a number of machine learning and data mining methods to accident data analysis, including methods such as the modularity-optimizing community detection algorithms, association rules learning algorithms, Bayesian Networks, and frequent pattern trees. The project has resulted so far in a Transportation Research Board (TRB) paper which was recently accepted for publication, and two other papers in refereed conference proceedings.

In this project, we propose a variational inference algorithm that tracks and predicts real-time traffic dynamics in a transportation network from an agent-based transportation model and multiple streaming data sources. To demonstrate the value of combining simulation modeling and big data in delivering travel information to drivers and promoting efficient driving, we will aggregate the noisy sensor network data from personal mobile phones on UB’s North Campus into transportation informatics such as available parking spaces in parking lots, and time and fuel consumption to find a parking space, suggest optimum trips to faculty, staff, and students through a real-time and interactive driving planner and promote fuel-efficient driving at UB’s North Campus. We will validate the variational inference algorithm and the driving planner with both synthesized and real data.

Performance measures and peer comparison are valuable tools that can be used as a step toward efficient and proactive management. These tools can be used to evaluate actual performance, identify areas that need improvement, and establish feasible goals for the near future.  With the most recent advances in technology it is possible to acquire a great amount of information with relative low costs or efforts. In the past, it was very difficult and costly to collect comprehensive performance data.

The overall goal of this study is to not only design a transportation data-warehouse prototype for the Buffalo-Niagara region, but also to demonstrate its usefulness through a specific application.

In the present day, data for vehicle accidents in Puerto Rico for the past five years are not available.  Graduate students from the University of Puerto Rico at Mayagüez Campus (UPRM) who wish to develop a research project regarding vehicle accidents cannot focus on data from Puerto Rico due to the lack of this type of information; these students have to rely on either out-of-date data or records from other countries.