There is a significant and growing need for travel time data. Transportation agencies need travel time data to support system operations and performance measurement. Information service providers need travel time data to provide a more competitive product. Given this need, numerous methods and technologies have been developed to estimate travel times. In some cases, these methods allow for in-house deployment of sensors and computing/communication infrastructure to allow an agency/company to produce the data that they will utilize. In other cases, companies have been founded that derive travel times from various sources (in some circumstances, the company's own sensor network, or from purchased probe location information) to produce a product - a travel time data service - that is then marketed to transportation agencies and information service providers.
Given this variety of approaches, and based on recent experience, it is clear that each source of travel time data offers distinct advantages and disadvantages. It is highly unlikely that a single source of data will emerge as a clear winner that dominates all other approaches. Because of this, consumers of travel time data (transportation agencies and information service providers) are faced with a need to choose one or more travel time data services to best meet their needs. To do so effectively, there is a need for a standard test procedure to use in assessing the quality of different travel time data services. This standard test procedure will create a level playing field that will allow consumers to comprehensively and fairly consider all travel time service options.
In today's environment, when a consumer of travel time data attempts to compare alternatives, the agency/firm is faced with a very confusing landscape that does not support fair and balanced decision-making. Most travel time data service firms will offer evaluations of their products. However, the only constant among these evaluations is that they are all different. The evaluation reports use different baseline (ground truth) data - ranging from point detectors to single or multiple probe vehicle runs, different statistical comparison methodologies - from rigorous hypothesis testing to very broad graphical approaches, and have been conducted by various parties, ranging from consultants to universities. Finally, in many cases, the evaluations were commissioned and paid-for by the data service providers themselves.
The result is that consumers of travel time data are forced to make multi-million dollar decisions based on a series of incompatible evaluation reports, many of which were not produced in an objective environment.
The objective of this project is to develop a standard test procedure to evaluate the quality of travel time data services. The standard test procedure will produce evaluation results that are consistent and will allow for fair comparisons between travel time data services.
Scope of Work
The research will produce an official standard sanctioned by a Standards Development Organization (SDO), such as ATSM, that will specify a clear quality assessment procedure for travel time data. This standard will then be available for transportation agencies and information service providers to specify when comparing methods/services. In other words, the standard test procedure will allow a travel time data services consumer to compare products of multiple methods/vendors on a level playing field.
To develop a standard test procedure for assessing travel time data quality, two parallel tracks must be coordinated at several key points:
1.Research on Test Procedures and Parameters. Given the large number of methods that have been used for travel time quality assessment, there is a need to objectively analyze these methods, in conjunction with a full understanding of consumer needs, to arrive upon a preferred travel time quality assessment method and associated parameters.
2.Standard Procedure Development. Once the preferred method and parameters has been identified, there is a need to identify the preferred SDO to sponsor the standard, and then a need to work through the standards adoption process.
See link to full study proposal for more details.
Suggested Contribution: $25,000/yr for 2 years