The Two-Way Street: People adapting to transportation and transportation adapting to people

Washington, DC
November 13, 2008

Poster Presentations

Usability and Accessibility Evaluation of Amtrak's Redesign of its Online Reservation System
Dustin Chambers, Angela Colter and Dick Horst ( UserWorks, Inc)
(Note: Two consultants played an integral role in the work presented here. Whitney Quesenbery worked with us on both the usability testing and accessibility review. Jennifer Sutton worked with us on the accessibility review)


Amtrak is in the process of redesigning its online reservation system. UserWorks has supported that process by providing design recommendations and usability and accessibility reviews of the emerging designs. We first usability tested the existing booking site with representative users and offered recommendations for the redesign. When the new design was in an early prototype stage, we provided an expert review by our usability specialists. More recently we conducted a usability test of a partially functional prototype and did an accessibility review of both the current site and redesign prototype. Our poster will illustrate the methods we have used. We will present the task scenarios used in the usability tests, the audience groups from which test participants were drawn, the usability measures being utilized, and an outline of the structure used for evaluating the accessibility of these web user interfaces. No specific findings will be discussed, but we can report that with the redesign there was higher success in completing basic tasks, more positive satisfaction ratings from participants, and fewer complaints about finding information needed to complete a booking.

Interactive Crash Analysis Application (EVA)
Darya Filippova (University of Maryland, Center for Advanced Transportation Technology Laboratory)

Accident analysis helps to identify the cause of the accident; analyzing multiple accidents helps to infer the high-accident locations, improve road safety, and potentially reduce accident severity. Mining accident data requires knowledge of querying languages and familiarity with the data. Currently, accident analysis tools are disjoint stand-alone applications thus slowing down the analysis. In particular, the tools used by agencies in Maryland are complex, cumbersome, and require training to perform even the simplest tasks. We propose EVA (Explore and Visualize Accidents) - a Web-based application that gives users the ability to query the accident data directly from any computer and analyze it through a series of visualizations. EVA's design can accommodate both computer novices as well as proficient users. EVA is unique in that it does not require the user to be familiar with the structure of the data to be able to explore it.

Transportation Data Analysis Web Tools
Jon Gilmour (University of Maryland, Center for Advanced Transportation Technology Laboratory) Real-time transportation operations data can come from detailed accident and construction reports, road sensors, weather systems, cameras, and automated vehicle positioning systems, and many other sources. While much of this transportation data is inherently spatial in nature, most archived transportation data systems display transportation data in only a temporal format-often in tables or simple graphs. The CATT Lab is investigating various techniques for returning and visualizing spatial transportation datasets. A map-based, historical transportation data query tool, circular time and space graphs, and other unique visualizations are being developed to enable transportation planners and researchers to gain new insights into complex temporal and spatial transportation data. This presentation will highlight some of the newest web-based tools being developed at the CATT Laboratory.


Usability Evaluation of E-Gov Travel Vendor Applications
Dick Horst (UserWorks, Inc.), Derek Schultz (Media Design Associates), Elizabeth Buie (Luminanze Consulting, LLC)
(Note: Several colleagues made important contributions to the work presented here. Our team included Andrew Schall, Claude Steinberg, Willie Stewart, Jenna Wexler, and an adviser, Ben Shneiderman.)


The government spends billions per year on employee travel expenses and is implementing an E-Gov Travel Solutions (ETS) program with the goals of reducing costs and increasing services. ETS does this by providing integrated, web-based applications for making travel reservations and managing the related planning, approval, and accounting processes. Recognizing that the usability of these ETS applications will be a major determining factor in government employees' use and satisfaction with them, the Program Management Office had UserWorks do a usability evaluation of the three vendors' systems. These evaluations involved expert reviews, analysis of user surveys, usability testing with government employees, and cognitive walkthroughs with "power users." A major challenge was the need to evaluate and report on the three systems' usability independently, without bias, and without compromising the competitive process among the three vendors. Our poster will describe how we organized our team, the methods we employed, and the lessons we learned, many of which will likely apply when introducing usability engineering into other large, complex, multi-vendor government programs.

Finding the Way Through Participatory Design
Adriana Machado, Michelle Robertson, Cynthia York (University of Baltimore School of Information Arts and Technology)

How can we encourage people to leave their cars at home and take advantage of public transportation systems? Instead of campaigning and trying to convince them by prescribing what's best for the environment, we might get better results by understanding people's needs and giving them tools to empower their informed decisions. Planning a trip from point A to B can be a very complex task, especially when it involves navigating diverse mass transit buses, trains, routes, schedules, and ticket options. Through Participatory Design sessions, we explored passengers' mental models and created concepts to assist their commuter trips. Facilitators and participants together designed and prototyped a mobile application that would not only help plan the trip but also serve as a "buddy" to count on in case unpredictable surprises happen along the way.

The Washington DC Regional Integrated Transportation Information System
Julian Moyse (University of Maryland, Center for Advanced Transportation Technology Laboratory)

The Regional Integrated Transportation Information System (RITIS) is an automated data sharing, dissemination, and archiving system initially designed for use in the National Capital Region. RITIS improves transportation efficiency, safety, and security through the integration of existing transit and transportation management data in Virginia, Maryland, and Washington D.C. The emphasis of RITIS is on data fusion and standardization, and their relationship to data collection, regional transportation systems management, regional traveler information dissemination, and system evaluation. RITIS automatically fuses, translates, and standardizes data obtained from multiple agencies in the region in order to provide an enhanced overall view of the region's transportation network. Participating agencies are able to view regional traffic information and use it to improve their operations and emergency preparedness. This presentation will focus on several traveler information components of RITIS including web-mapping, alert systems, 511, and mobile web applications.

Configurable Transportation Event Management System
Andreea Olea (University of Maryland, Center for Advanced Transportation Technology Laboratory)

Effective transportation incident and construction management in major metropolitan areas is critical to the safety and reliability of transportation operations. Efficient traffic management can save heavily populated regions billions of dollars in lost wages, time, fuel consumption, and environmental pollution. Numerous transportation departments have created their own proprietary software systems to help better manage the roadways. These systems, however, are expensive to build and often very difficult to use. This project strives to develop a user-friendly, intuitive, and feature rich transportation management system that can be easily configurable by any transportation department looking to start a transportation management operation. Our system offers a customizable web-based interface to facilitate communication and coordination critical to event management and incident response.

I-95 Corridor Coalition Mile-marker & Ramp Designation Sign Study
Michael L. Pack (University of Maryland, Center for Advanced Transportation Technology Laboratory)

Mile marker and ramp signs are critical informational components for motorists, transportation system managers, maintenance, law enforcement, fire, and emergency responders for reporting of and responding to collisions or other incidents. Lives have been saved or lost because of the ability or inability of a motorist to accurately read and understand a mile marker or ramp designation sign to identify their location. Misconstrued signing can delay response to incidents, jeopardizing lives and compromising the effective and efficient use of resources. Providing crash location information becomes more difficult on complex urban highway interchanges. Inadequate ramp designation signing may lead to incorrect locations being given to 911 centers. The goal of the research performed in this study was to develop better mile marker and ramp designation signs using (1) a computer model of human nighttime legibility (2) a computer-based sign comprehension analysis, and (3) a road test validation.

Considerations for Using Eye Trackers During Usability Studies of Signs
Anjali Phukan (University of Maryland Baltimore County)

The purpose of this usability study was to see if eye trackers really collect valid data, regardless of the user's method of corrected vision, eye color, or gender. The study also explored ways the user's first language can affect the way the user gazes (and hence interprets) text and images as part of a look into the international need for effective signs, regardless of a user's origin. This is all within the context of usability testing for signs used in transportation systems that must accommodate global users, including a wide variety of users who may not have the same method of vision correction, or share the same first language. The motivation to explore the idea that these human factors can distort the eye tracker is based on marketing claims by several companies that say these factors should not effect the results. This study found that the validity of data in usability studies that involved eye trackers in testing can produce biased results based on eye wear and eye color, and that adjustments should be made to the study to control for these variables. The results showed no significant correlations based on gender. In addition, there are indications that the alphabet of a user's first language may have some impact when reading and viewing English text directions or a Latin alphabet based visual dictionary.

Travel TAMEr: PDA Interface for Travelers
Tami Smith, Amy Rubino, Musa Karakaya, Emily Marsh (University of Baltimore School of Information Arts and Technology)

The tourism field presents a perfect domain for a ubiquitous computing application (Weiser, 1991). Travelers have complex needs: locating, navigating, searching, identifying, and checking (Reichenbacher, 2001) a broad range of resources which need to be met in complex, rapidly changing settings. Consequently, PDA-sized tools are becoming a more convenient way for travelers to navigate public transportation. This project created such a tool based on a foundation of ethnographic research, participatory design, and technological functionality. It is designed to meet travelers' needs for novelty, immersion, and safety both specific to transportation and in other contexts.

The conceptual basis for the tool rests upon three original user personas based on ethnographic research (Wickens, 2002; Yiannakis & Gibson, 1992): the tourist, the explorer, and the wanderer. The other features of the Travel TAMEr arose out of participatory design sessions with potential users-minimal clicking, feedforward features, and providing a unique experience.

Examining Child Seat Installation Errors of Novices & Parents
Yi-Fang D. Tsai (George Mason University, NHTSA)

Child restraint systems are known to be complicated to install and can be affected by factors such as features in the child seat design, vehicle design, and choosing which attachment system to use. Past studies have evaluated misuse patterns at car seat checkup locations where experience handling a child safety seat may vary. Two studies were conducted to examine child seat misuse with samples of varying child seat experience. The first study observed novices as they installed a convertible child seat. The second study was conducted with current child seat users where their child seat was inspected for a correct installation. After the assessment of errors in both studies, participants were interviewed on their experience with child seats and asked to rate options from a booklet of usability improvements.

Real-time and Historic Incident Visualization Using Timelines
Michael VanDaniker (University of Maryland, Center for Advanced Transportation Technology Laboratory)

Traffic management centers worldwide use software applications to record and coordinate incident response. Proper decision making depends on the incident manager's ability to understand all of the data gathered by these tools, but existing visualizations have failed to make this an easy task. Currently, managers must analyze several pages of free text, tables, and maps to understand what is being done to manage an incident. This can be time consuming, and the consequences of misinterpreting information can be life threatening.

This project created a compact, one-screen visualization of the temporal and spatial data associated with traffic incidents. Temporal data are projected onto a timeline representing the entire incident. A built-in map allows the user to garner spatial significance while still providing temporal information. This tool reduces the chance of missing critical information and reduces the time needed to comprehend the numerous events that occur during the course of an incident.

4D, Real-time Transportation System Visualization
Phillip Weisberg (University of Maryland, Center for Advanced Transportation Technology Laboratory) The 4-Dimensional, wide area traffic visualization tool provides incident management personnel, emergency management personnel, and the general public with the means to achieve situational awareness in as close to a real-world setting as possible. This research developed a system for visualizing 4D, (where the fourth dimension is time), "real-time" transportation data. This effort employed a combination of technologies including OpenGL and various modeling techniques to develop a scalable, highly interactive 4D model using available GIS and transportation infrastructure data in conjunction with real-time traffic management center data. The prototype system interacts with real-time traffic databases to show animations of real-time traffic, incident, and weather data. A user can "fly" or "drive" through the region to inspect conditions at an infinite number of angles and distances. The program also allows users to monitor the status of and interact with traffic control devices such as dynamic message signs, CCTV feeds, traffic sensors, and even view the location of a fleet of vehicles equipped with GPS transponders.

Visual Analytics for Transportation Incident Datasets
Krist Wongsuphasawat ((University of Maryland, Center for Advanced Transportation Technology Laboratory)

Transportation systems are being monitored at an unprecedented scope which is resulting in tremendously detailed databases. While the transportation community emphasizes developing data standards, little effort has been made to design appropriate visual analytics tools to explore the data, extract meaningful knowledge, and represent results. Analyzing these large multivariate geospatial datasets is a non-trivial task.

A tool that affords sophisticated yet user-friendly analysis of transportation incident datasets is proposed. Interactive maps, histograms, two-dimensional plots and parallel coordinates plots are four featured visualizations that are integrated together to allow users to interact with and see relationships between multiple visualizations. With a rich set of filters, users can create custom conditions to filter data and focus on a smaller dataset. Since finding interesting relationships can be time-consuming due to the multivariate nature of the data, a rank-by-feature framework has been adopted to quantify the strength of relationships between the different data fields.