1 Development of a Traffic Information System Driver Interface

Based on ideas from the literature and the project team, several reasonable interface options were prototyped using SuperCard (on a Macintosh computer). Some designs were eliminated after initial prototype attempts because critical details were not legible. Others were developed further and evaluated using human factors computational methods (GOMS, Tullis Display Analysis Program) and quick usability tests. Five basic designs have been developed. For the designs where simple text displays show incidents and congestion, there are three ways to select the highway of interest: a bidirectional scrolling menu, a graphic of the highways, and a phone-style keypad. The graphic designs use a skeleton map of the metro Detroit highways, and represent traffic problems through color coding or explicit travel speeds. An experiment was conducted at a motor vehicle licensing office to examine the understandability of and preferences for aspects of the five designs. From this it was determined that using orange to indicate slow moving traffic is ambiguous to drivers. Drivers preferred a color coding scheme which used green to indicate freely moving traffic (at or above speed limit), yellow to indicate slow traffic (30 54 mph), and red for stop and go traffic (0 29 mph). The results of this study were used to modify the interfaces for future testing in a driving simulator. Introduction The annual cost of traffic congestion in the United States is believed to be as much as 61 billion dollars [1, 2]. The implementation of Intelligent Vehicle-Highway Systems (IVHS) should help reduce this waste of money and the associated waste of time. One way to accomplish this is to provide motorists with current traffic information throughout their trip. Given the proper information, many drivers may be willing to modify their routes or departure times to avoid delays due to congestion or incidents [3]. Motorist information surveys have shown that the most important factor in route choice for daily commuters is travel time [3, 4, 5]. For the average daily commute of 30-35 minutes, a delay of 16 18 minutes [3, 6] would cause drivers to divert to an alternate route. Although the potential benefits of providing traffic information have been established, the most effective and safe methods for providing the information to drivers have not. For a traffic information system to be effective, it must be easy for drivers to understand and use while driving. To address usability concerns, traffic information system interfaces were developed using rapid prototyping software and evaluated by potential users. This effort is part of a large project funded by the U.S. Department of Transportation to develop human factors guidelines and methods for testing the safety and ease of use of future driver information systems. A model predicting human performance with these systems is also being produced. An overview of the project is described in [7]. This project has several unique aspects. First has been the application of the iterative design approach [8] commonly used for user-computer interface development, to automotive design. Second, analytic methods used to evaluate user-computer interfaces such as GOMS and the Tullis Display Analysis Program have also been applied for the first time in the automotive domain.