Focus on a basic set of query and display capabilities sought by the majority of users when designing an Archived Data Management System (ADMS).
Lesson learned from Virginia's experience evaluating usage of a ADMS.
Made Public Date
03/26/2014

966

Virginia
United States
TwitterLinkedInFacebook
Identifier
2014-00677

Usage Analysis of a First Generation ITS Data Archive: Lessons Learned in the Development of a Novel Information Technology Application

Background

Transportation professionals have been exploring the uses of archived data from real-time operations such as traffic control and toll collection to support offline activities such as operations planning, long-range planning, and performance measurement. A systematic approach to the use of archived data is fully defined in the Archived Data User Service (ADUS) within the National ITS Architecture. The actual deployment of an ADUS is called an Archived Data Management System (ADMS).

The Commonwealth of Virginia conducted a federally sponsored field operational test to develop a first generation ADMS. The resulting product, ADMS Virginia, is a web-based software application to support archived traffic data retrieval and access to a suite of decision support tools built upon the already established traffic data archive created and maintained at the Virginia Smart Travel Laboratory. The Smart Travel Laboratory (STL), a joint facility of the Virginia Department of Transportation (VDOT) and University of Virginia (UVA), is the official archive for all ITS data in the Commonwealth of Virginia.

The team used a spiral development approach, as opposed to the waterfall approach more commonly used in transportation applications, to allow users to see and work with a number of intermediate products, providing several opportunities for substantive input. Key stakeholders were identified during the proposal stage, and were included in making all major decisions throughout the project. The users were also presented with quick prototypes before the start of each of the four builds, to provide feedback based on their reactions. After the release of each build, users were surveyed to get specific post-deployment feedback. As of July 1, 2005, there were a total of 164 ADMS Virginia users representing 61 different organizations.

Lessons Learned

Based on a quantitative analysis of users and uses of the system, as well as user feedback, the following lessons were learned:

  1. Don't overspecialize early in a system's life. Usage trends of ADMS Virginia made it clear that most users wanted access to basic data query and display capabilities. Eighty percent of the use of the system was devoted to downloading data and creating simple plots and/or maps based on this data. Specialized services that took considerable design and development effort were lightly used.
  2. Anticipate a wide range of users. Demand for traffic data is great, and comes from a diverse group of people, including the research community. Allowing for wide access will minimize time spent allowing "special" access for groups that were not originally expected to use the system.
  3. Avoid spiral development as it creates the risk of establishing system usage patterns prematurely. Experience in ADMS Virginia showed that users tend to stick with usage patterns early, even failing to utilize design requested improvements in later builds. For example, mapping services were added in Build 4. The user group that received this build as their first interaction with the system tended to use this feature more than the group that had this feature added to the system they were already familiar with.

The lessons learned during the ADMS Virginia development effort are of value to individuals seeking to create an Archived Data User Service (ADUS) as defined within the National ITS Architecture, to those seeking to make use of archived ITS data, and to those involved with creating more general classes of new information technology systems for transportation applications.