ABSTRACT

The current design process for vibratory alerts relies heavily on physical prototyping and human-subjects testing. Models of attention could be used to help guide the early design stages before testing begins. Such models exist for visual attention but not for the tactile modality. To begin addressing this gap, the Noticing-Saliency-Effort-Expectancy-Value (NSEEV) model was used to predict the time to first fixation and response accuracy of alerts in a medium-fidelity driving simulator with a main following task and two types of alerting events: Blindspot detection and forward collision avoidance. Vibrations were 30 participants were recruited for the study and compensated for their time. Preliminary results of the study and the application of the model are to be discussed. Response time, response accuracy, effects of SEEV variables on those performance metrics, and a comparison between empirical results and the model output are all being considered. The goal of this study is to determine whether this application of the NSEEV model is suitable for designers to use when initially checking the feasibility of their vibrotactile alerts’ parameters.

BIOGRAPHY

James Parkinson is a PhD. Candidate in the Edwardson School of Industrial Engineering at Purdue University. He is co-advised by Dr. Brandon Pitts (NHanCE Lab at Purdue) and Dr. Steve Landry (Penn State University). He received his B.S. in Industrial Engineering from Purdue in 2016. His research is focused on ways to improve alerts so that they better support users in their unique contexts. His dissertation focuses on vibrotactile alerts, especially when applied in vehicles. He is also interested in their use with wearable devices like behavior change support systems.