 Information systems are becoming more ubiquitous in everyday life, and the automobile is a promising location for new mobile devices. However, systems in cars are subject to strong constraints even beyond those typical for mobile devices. Designing systems to satisfy these constraints requires an understanding of many aspects of the domain, including the physical and cognitive demands of driving.
Our research in this area has three foci:
- understanding how drivers manage the demands of both driving and operating an in-vehicle information system,
- user testing of in-vehicle information systems in a low-fidelity driving simulator, and
- computational cognitive modeling of driving and information-system use.
|
We have begun a literature review to gather information from basic and applied psychology about various aspects of managing several simultaneous tasks. We are pulling from areas as diverse as interruptions, dual-task, and task switching. Information from this literature review will serve as inspiration for design ideas and principles and as input into our modeling work. The papers we reviewed and our preliminary analysis can be found under the Publications tab above. |
We have constructed a low-fidelity driving simulator to collect human data on in-vehicle information systems. Design ideas can be mocked-up on a heads-up display or a touch screen display mounted to the right of the steering wheel. We also plan to compare data collected on the simulator to the predictions of our computational cognitive models to inform our modeling effort.
This simulator has been used by student groups to test their ideas (see the Connections tab above).
A poster about our simulator can be found at the Publications tab above. |
 |
 |
We have joined with Dr. Dario Salvucci at Drexel University to make it easy to create computational cognitive modeling of driver performance with in-vehicle information systems. An analyst can mock-up an information system in html, demonstrate a task using that mockup and automatically integrate a model of that task with a model of driving. The resulting model produces quantitative predictions of measures of interest such as lateral deviation from the center of the road and time to perform the information task, as well as a timeline of actions such as hand and eye movements.
A poster about our modeling effort can be found at the Publications tab above. |