Automotive User Interfaces and Interactive Vehicular Applications

The Process of Creating an IVAT Plug-in: Mirror Checking
Thomas M. Gable, Richard Swette, Alan M. Tipert, Bruce N. Walker & Sundararajan Sarangan
Sonification Lab, School of Psychology
Georgia Institute of Technology
654 Cherry Street Atlanta GA 30332 USA
thomas.gable@psych.gatech.edu, {rswette3,tipert,bruce.walker,sundar}@gatech.edu
ABSTRACT
In this paper, we discuss the process of creating a mirror checking
plug-in for a personalized, adaptive program used to assist drivers
who have had a traumatic brain injury (In-Vehicle Assistive
Technology, or IVAT). Our aim in the current paper is to outline
the procedure utilized in the creation of the current IVAT plug-in
and to highlight the importance of user-centered design in an
application such as this. We discuss the myriad of decision points
and design possibilities and how we base implementation in
theory about user needs and capabilities.
Categories and Subject Descriptors
H.5.2. [Information Interfaces and Presentation (e.g., HCI)]:
User Interfaces – interaction styles (e.g., commands, menus,
forms, direct manipulation), user-centered design
General Terms
Design, Human Factors.
Keywords
User Interface, Traumatic Brain Injury, Assistive Technology,
Human Factors, Plug-in Design
1. INTRODUCTION
Each year 1.5 million Americans report new brain injuries in
addition to the 5.3 million already identified as having a traumatic
brain injury (TBI) [1]. While the injuries significantly impact their
lives, these individuals often attempt to become fully independent
post-injury, including undertaking independent driving [2]. The
cognitively and perceptually demanding task of driving, however,
can create issues for individuals with a TBI due to the consequent
impairment in cognitive functioning it presents [2]. Individuals
who have experienced TBIs have often reported problems with
safe driving practices, creating situations in which driving can be
dangerous [3].
2. IVAT Background
To assist with the re-integration of individuals into mainstream
society following TBIs, the Sonification Laboratory at the
Georgia Institute of Technology has been working with the
Shepherd Center to create and implement an In-Vehicle Assistive
Technology (IVAT) system. Rehabilitation experts at the
Shepherd Center originally developed a hardware device (a
“three-button box”), called the electronic driving coach (EDC),
for a case study of an individual with a TBI [4]. The Sonficiation
Lab has extended and expanded this proof-of-concept into a
complete software system that can now be installed onto a car PC
integrated directly into the vehicle [5]. From the outset researchers
sought to understand user needs for the global IVAT system
through interviews and focus groups, as well as iterative
�����������������������������
�����������������������������������������������������������������
�������������������
- 27 -
evaluations of prototypes with potential users and driving
rehabilitators [5]. That research identified a set of tasks performed
while driving that could be affected by a TBI. The vision of the
future IVAT system is a system of plug-ins, each designed
specifically for one of these tasks. These plug-ins could be
employed alone or in combination with others, ultimately creating
a personalized and adaptive driving system for any given user. It
is now important to ensure that the development and design of
each one of these IVAT plug-ins is appropriate and effective.
Drivers with TBIs present unique challenges in developing such
an automotive user interface.
3. The Plug-in: Mirror Checking
3.1 Choosing Mirror Checking
Considering driving as a whole, maintaining spatial and
situational awareness (SA) is a vital part of safe driving [6]. Given
the high priority of and TBI drivers’ frequent problems with SA,
facilitating the task of checking mirrors was selected to be the
function of the first purpose-built IVAT plug-in (see e.g., [7] for
the empirical research that assesses driver situation awareness
using mirror checking).
3.2 Primary Considerations
During the design process the overall goal of a plug-in must be
well established. In the case of this mirror checking application,
the goal was to increase the users’ awareness of their surroundings
by reminding them to check their mirrors, making them safer and
more engaged drivers. The system must accomplish this goal
without adding significantly to the user’s cognitive load and
inadvertently impairing his or her ability to drive. Additionally,
the plug-in must be compelling enough to keep the user’s
attention, which has been shown to be important to drivers with a
TBI (because drivers with a TBI frequently forget that they are
driving) [5]. Furthermore, the system must unify the mirrorchecking
task with driving, and not create a situation with dual
tasks.
3.3 Planning the System
The system has two basic functions: issuing a reminder to
complete a task and registering completion of the task. The first
consists both of determining appropriate times to issue reminders
and of the reminder itself. The second also requires two steps,
appropriately recognizing completion and communicating that
recognition. By separating these steps, each variable can be
considered in stride during development and receive either a
design decision based on previous knowledge and research or an
empirical investigation.
3.4 Making Design Decisions
An early decision was made to exploit the utility of multimodal
interfaces to maximize the effectiveness of a notification while
minimizing its additional cognitive load. This practice is
supported by multiple resources theory and discussed within the
context of the IVAT project in concurrent research [8,9]. During
the development of the mirror checking plug-in, researchers were