An Investigation of Usability Issues with Mobile Systems Using a Mobile Eye Tracker

1
An Investigation of Usability Issues with Mobile
Systems Using a Mobile Eye Tracker

• thesis by
• Marie Norlien
• International University in Germany
• Thesis supervisors
• Prof. Dr. Anthony Jameson Prof. Dr. Karl Rohr

2
Outline

• Introduction
• Mobile eye tracker
• Study performed of mobile systems with a mobile
  eye tracker
• Conclusion
• What is left to do
• Acknowledgements

3
Introduction

• Mobile systems raise many usability issues that
  are not yet well understood.
• One technique for acquiring data relevant to
  these issues is the use of a head-mounted mobile
  eye tracker
• The goals of this thesis are
• To explore the potential and limitations of an
  eye tracker in example studies of one or more
  typical mobile systems
• To contribute some insights into particular
  usability problems that arise with those systems.

4
Outline

• Introduction
• Mobile eye tracker
• Study performed of mobile systems with a mobile
  eye tracker
• Conclusion
• What is left to do
• Acknowledgements

5
Eye tracker

• ASL Model 501 mobile eye tracker
• Worn on users body
• One camera records the scene the user is looking
  at the other camera records the users eye
  movements

6
Eye tracker

• Strengths
• Subject is free to move without losing accuracy
  of calibration
• Mobile, within the range of the video receiver
• Subject can wear glasses
• Limitations
• Subject is wearing many pieces of equipment,
  which can be uncomfortable
• Difficult to calibrate and set up initially
• Viewing plane is restricted to the view of the
  scene camera
• Cannot use outside

7
Eye tracker

• Calibration process
• Subjects must be in a fixed position during
  calibration for the best recording of accurate
  data
• The subjects eye must not be obstructed by the
  subjects eyelid
• For example, the eyelid mostly covers the eye
  when a subject is looking down at a Pocket PC

8
Eye tracker

• Potential barriers to accurate calibration
• Physical movement of the subject
• Even slight movement unnoticed by the subject
• Excessive overhead light, sunlight, or spot
  lighting
• Inadequate pupil and/or corneal reflection
• Solutions
• Good Calibration methodology
• Subject should be fixed during calibration
• Preventing light with the use of a sun-umbrella,
  dark walls, and anti-glare calibration surfaces

9
Outline

• Introduction
• Mobile eye tracker
• Study performed of mobile systems with a mobile
  eye tracker
• Conclusion
• What is left to do
• Acknowledgements

10
Study performed

• We analyzed two types of systems
• Xybernaut a small handheld PC with a basic
  indoor navigation system based on infrared
  sensors
• REAL software developed by the Department of
  Computer Science, Saarbrücken University
• The Compaq Ipaq or Hewlett-Packard (HP) Jornada
  pocket PC installed with market Global
  Positioning System (GPS) hardware and software

11
Study performed

• Xybernaut
• Functionality includes indoor navigation via maps
  visible to the user from the system screen
• The system is configured for a planned route in
  the Computer Science department at Saarbrücken
  University
• Investigate how a user can manage multi-tasking
  with a mobile system

12
Study performed

• Xybernaut

Xybernaut hardware with REAL software
13
Study performed

• Pocket PC
• Functionality includes outdoor navigation via a
  map visible to the user from the pocket PC
  screen, which the user can manipulate via menus
  and icons
• Compared the functionality of two already
  existing market GPS navigation systems
• World Navigator by Teletype GPS
• Destinator Personal Navigation System by RACO
  Industries
• Investigate how user-friendly and intuitive each
  interface is

14
Study performed

• Pocket PC

HP Jornada hardware with Teletype software
Compaq Ipaq hardware with Destinator software
15
Study performed

• Results analyzed
• Interview with subject while viewing each video
  recording
• Targeted Questions
• Event Tables
• From the videos, each event was logged
  systematically (begin end time)
• PERT diagrams
• From videos and Event Tables the following were
  derived
• 3 diagrams to describe general case behaviors
• 2 diagrams to describe multi-tasking
• Pocket PC evaluation

16
Study performed

• Event table example

17
Study performed

• Event Tables Summary
• Main difference in subjects
• Event-driven
• Interested in starting to perform the task
• Once there was an event, the user adjusted their
  plan to solve the task
• Plan-driven
• Developing a strategy for each task through map
  dragging
• Having an understanding how they are positioned
  on the map relative to the environment

18
Study performed

• PERT diagram example

19
Study performed

• PERT diagrams summary
• 3 possible answers why users perform tasks
  sequentially vs. Multi-tasking
• The user may find it too hard to perform the
  actions with the system
• The user finds the task so demanding cognitively
  that the user must focus on one behavior at a
  time in order to manage the demands
• The user needs further information from the
  system before it makes sense to move forward.

20
Study performed

• Pocket PC summary
• How the menus should appear and work (4
  Usability principles Dix, et al., 1997)
• Familiarity
• Generalizability
• Predictability
• Task Migratability
• General comments regarding problems encountered
  while walking

21
Outline

• Introduction
• Mobile eye tracker
• Study performed of mobile systems with a mobile
  eye tracker
• Conclusion
• What is left to do
• Acknowledgements

22
Conclusion

• To understand the concepts of multitasking
  applied to a mobile system
• There are three basic guidelines
• Consider possible uses in the environment of the
  system in combination with various user behaviors
• Consider including the possibility of
  extraordinary events occurring with the system
• Consider situations were the user should not
  focus their attention on using the system

Jameson, A. (2002). User Multitasking as a Design
Challenge for Mobile Multimodal Systems. ITRW on
Multi-Modal Dialogue in Mobile Environments
Kloster Irsee, 17-19 June 2002
23
Conclusion

• Eye tracking can provide additional valuable
  information to a well designed usability study in
  various structured environments leading to
  important usability discoveries regarding a given
  mobile system
• Despite its limitations
• User interfaces PocketPC Navigation Software
• Identifies the complexity of user Multi-tasking
  while using a mobile system

24
Conclusion

• What is left to do
• Find the link to Multi-modal systems
• Struck from thesis title
• Multi-tasking?

25
Conclusion

• Acknowledgements
• Thank you to everyone, it would not have been
  possible without you.
• Prof. Dr. Anthony Jameson, DFKI GmbH (German
  Research Center for Artificial Intelligence)
• Prof. Dr. Karl Rohr, International University in
  Germany
• Kerstin Klöckner, DFKI GmbH, Researcher
• The department of computer science, Saarbrücken
  University.
• Christoph Stahl, Rainer Wasinger, and Christian
  Müller

26
Questions ?