APPLYING SPATIAL ABILITY ASSESSMENTS TO A PERSONALIZED NAVIGATION ASSISTANT

S. Fickas1, X. Yao1, A.K. Lobben2

1 - University of Oregon, Department of Computer and Information Science

2 - University of Oregon, Department of Geography

fickas@cs.uoregon.edu

 

This paper hypothesizes that a navigation-assistance system (e.g., OnStar) that is personalized to a single users goals, abilities, needs and preferences, can lead to better results than a non-personalized generic version. An experiment is proposed that uses individual performance on spatial ability tests as the basis for personalization. Each subject will be assessed on their map rotation abilities. Previous research shows that map rotation is a task involved in navigational map reading. When traveling in a direction (other than a straight line on a map) a person is faced with the rotation task and must choose between one of two strategies: physically rotate the map or mentally rotate the map geometry and accompanying text. Literature also documents the effects of rotation angle on a persons ability to effectively complete the rotation task. Therefore, map rotation may not be measured in only binary terms (i.e. people can or can not rotate maps), but measured along a continuum, where different people are affected by different angles/rotation thresholds.

Our PNA is part of an ongoing project that supports pedestrian travel in a city environment. In particular, the SmartBag system has been developed in the Wearable Computing Lab at the University of Oregon. The bag itself has onboard GPS for determining location and a magnetometer for determining heading. The bag also holds a PDA that provides visual directions. The bag communicates wirelessly to the PDA to provide location and heading; the PDA uses that to present directions to the user. The experiment proposed will vary the actual presentation in terms of rotation of the image. Before a subject is given the bag/PDA system, they will be assessed based on pre-test performance of map rotation. Given the results, a parameter will be set in the system that controls rotation. This parameter will be individual/specific to each subject, based on the pre-test thresholds that identify the angles at which subjects can no longer effectively perform map rotation. The PNA and subject performance will then be tested in a city environment.

The study proposed here is the first in a series of interest. Beyond map rotation, other personalization points can be taken from spatial abilities including visual memory and self-location abilities. We hypothesize, based on previous research, that each of these have the potential to personalize a navigation system to the greatest benefit to an individual.