IWAAL 2013 - Mobile NFC vs Touchscreen Based Interaction: Architecture Proposal and Evaluation
1. Mobile NFC vs Touchscreen Based Interaction:
Architecture Proposal and Evaluation
IWAAL 2013
Pablo Curiel, Koldo Zabaleta, Ana B. Lago
DeustoTech - Deusto Institute of Technology, University of Deusto
http://www.morelab.deusto.es
December 3, 2013
NFC vs Touchscreen Based Interaction
1/38
4. Advances in ICTs
►
Increasing number of electronic devices with noticeable
computing capabilities.
NFC vs Touchscreen Based Interaction
Introduction
4/38
5. Advances in ICTs
►
►
Increasing number of electronic devices with noticeable
computing capabilities.
Wide range of advanced services offered.
NFC vs Touchscreen Based Interaction
Introduction
5/38
10. Related Projects
►
►
►
Several studies have used NFC-enabled phones to ease
the interaction with them and provide different services.
Common approach: Each NFC tag identifies a given
service that is fired when the tag is read.
Extensions to it:
► General
tags which identify objects and special tags with
additional information or services that those objects provide
(Riekki et al.).
► User context as additional condition to determine which
service to provide (Riekki et al.).
► Tags that identify services and tags that identify arguments
for them. (Broll et al.).
NFC vs Touchscreen Based Interaction
State of the Art
10/38
11. Our Proposal
A tag represents an object or concept of the
real world, not a particular service. Thus, it is
the combination of tags read by a user what
determines the service to be provided.
NFC vs Touchscreen Based Interaction
State of the Art
11/38
13. The Platform
►
Activate the most used services on mobile devices by
interacting with NFC tags.
►
Client-Server architecture
►
Three main components:
► The
Model
► The Server
► The Mobile Application
NFC vs Touchscreen Based Interaction
Proposed Solution
13/38
14. The Model
►
OWL Ontology where two kinds of elements are modelled.
NFC vs Touchscreen Based Interaction
Proposed Solution
14/38
15. The Model
►
OWL Ontology where two kinds of elements are modelled.
► Real
world entities to be represented on NFC tags
NFC vs Touchscreen Based Interaction
Proposed Solution
15/38
16. The Model
►
OWL Ontology where two kinds of elements are modelled.
► Real
world entities to be represented on NFC tags
► Hierarchy
NFC vs Touchscreen Based Interaction
Proposed Solution
16/38
17. The Model
►
OWL Ontology where two kinds of elements are modelled.
► Real
world entities to be represented on NFC tags
► Hierarchy
► The
actions to execute
NFC vs Touchscreen Based Interaction
Proposed Solution
17/38
18. The Model
►
OWL Ontology where two kinds of elements are modelled.
► Real
world entities to be represented on NFC tags
► Hierarchy
► The
actions to execute
►A
combination of different number and types of tags
hasTag exactly 1
EmailTag
EmailTag
SendEmail
hasAttribute min 1
Contact
NFC vs Touchscreen Based Interaction
Proposed Solution
Contact
18/38
19. The Server
►
Checks if the combination of tags read by the user is
valid…
And if so, it determines the action they represent.
►
Two steps:
►
► The
Rule Engine
► Consistency Checking
NFC vs Touchscreen Based Interaction
Proposed Solution
19/38
20. The Server – The Rule Engine
One rule per action or service available.
► Each rule checks if the information of the tags read
matches, in principle, to its corresponding action.
► If so, it creates an instance of that action and with the
attributes that apply in each case.
►
Telephone
And
Contact
instances
NFC vs Touchscreen Based Interaction
Rule
Engine
Proposed Solution
Telephone
action
with associated
Contact
20/38
21. The Server –Consistency Checking
►
Once an action with its corresponding attributes is
created, its consistency is checked.
► Types
of the attributes
► Cardinality of the attributes
TelephoneCall
Telephone
action
with associated
Contact
Semantic
Reasoner
hasAttribute
exactly 1
Contact
Contact
NFC vs Touchscreen Based Interaction
Proposed Solution
21/38
22. The Server –Consistency Checking
►
Once an action with its corresponding attributes is
created, its consistency is checked.
► Types
of the attributes
► Cardinality of the attributes
►
Consistent Action = Valid Action
TelephoneCall
Telephone
action
with associated
Contact
Semantic
Reasoner
Consistent!
NFC vs Touchscreen Based Interaction
hasAttribute
exactly 1
Contact
Contact
Proposed Solution
22/38
23. The Mobile Application – Object Creation
►
The representation of the objects in the tags follows the
ontological model.
NFC vs Touchscreen Based Interaction
Proposed Solution
23/38
24. The Mobile Application – Object Creation
►
The representation of the objects in the tags follows the
ontological model.
<PlaceTagUri>
NFC vs Touchscreen Based Interaction
Proposed Solution
24/38
25. The Mobile Application – Object Creation
►
The representation of the objects in the tags follows the
ontological model.
<PlaceTagUri> <rdf:type> <lnfc:PlaceTag> .
<PlaceTagUri> <lnfc:latitude> “51.513016”^^xsd:double .
<PlaceTagUri> <lnfc:longitude> “-0.122337”^^xsd:double .
NFC vs Touchscreen Based Interaction
Proposed Solution
25/38
26. The Mobile Application – Action Execution
►
When a user reads a combination of tags to execute an
action, the URIs written on them are used to retrieve the
instance from the triplestore.
►
The retrieved instances are sent to the server, which
determines the action to execute.
►
Finally, the mobile phone provides the service that
corresponds according to the action returned by the
server.
NFC vs Touchscreen Based Interaction
Proposed Solution
26/38
28. Evaluation
►
Goals
► Validate
our proposal
► Compare it to traditional Touchscreen-based interaction
►
Hypothesis: Interaction based on NFC is faster and
easier for end-users than the touchscreen- based one.
NFC vs Touchscreen Based Interaction
Evaluation
28/38
29. Experiment design
►
Subjects executed three different actions, both with NFC
and an ad-hoc touchscreen-based GUI.
► See
weather forecast (no params)
► Make a phone call to a contact (one param)
► Send an e-mail to two contacts (two params)
NFC vs Touchscreen Based Interaction
Evaluation
29/38
33. Experiment steps
Application explanation
► Live demo
► Training for one subject group
► Subjects executed the actions
►
► Half
of the subjects first with NFC, the other first with
touchscreen
►
Post-experiment survey
► Age,
Gender, Familiarity with NFC/Touchscreen
phones/New technologies in general and subjective
impressions
NFC vs Touchscreen Based Interaction
Evaluation
33/38
35. Subject characteristics
40 Subjects
► Between 20 and 60 years old
► 30 men and 10 women
► Varied technological skills
►
NFC vs Touchscreen Based Interaction
Evaluation
35/38
36. Subject characteristics
40 Subjects
► Between 20 and 60 years old
► 30 men and 10 women
► Varied technological skills
► 80% owned a touchscreen phone
► 66% familiar with NFC/RFID
►
► But
►
only 33% had used a NFC-enabled phone
15 subjects trained with the application before the
experiment
NFC vs Touchscreen Based Interaction
Evaluation
36/38
37. Subject characteristics - Dependencies
►
Age and technological skills
► Older
subjects took less advantage of their smartphones
NFC vs Touchscreen Based Interaction
Evaluation
37/38
38. Subject characteristics - Dependencies
►
Age and technological skills
► Older
►
subjects took less advantage of their smartphones
Age and phone type
► All
subjects under 40 owned a smartphone
NFC vs Touchscreen Based Interaction
Evaluation
38/38
39. Subject characteristics - Dependencies
►
Age and technological skills
► Older
►
Age and phone type
► All
►
subjects under 40 owned a smartphone
Age and experience with NFC
► None
►
subjects took less advantage of their smartphones
of the older subjects had used a NFC phone before
Phone type and experience with NFC
► Only
one subject with prior experience with a NFC phone
owned a non-touchscreen phone
NFC vs Touchscreen Based Interaction
Evaluation
39/38
40. Subject impressions
►
►
Nearly all subjects found the NFC-based interaction
useful
75% of the subjects would use it at least occasionally
NFC vs Touchscreen Based Interaction
Evaluation
40/38
41. Task efficiency
►
NFC 0.5 seconds faster on average
►
No significant difference between the different actions
21 subjects faster with NFC / 19 faster with touchscreen
Pre-experiment training equal for both interaction
approaches
►
►
► Subjects
with training were 2 seconds faster on average
with both technologies.
►
No correlation between subjective impressions and task
efficiency
NFC vs Touchscreen Based Interaction
Evaluation
41/38
42. Task efficiency (II)
►
Significant correlation between task efficiency and
subject age
► Specially
► 630
ms slower per year increase (P < 0.001)
► Weaker
► 333
for touch-based interaction
for NFC
ms slower per year increase (P < 0.01)
NFC vs Touchscreen Based Interaction
Evaluation
42/38
43. Task efficiency (III)
►
Interaction time differences (touchscreen time minus
NFC time for each subject and task)
► For
each year increase in subject age, they are 300 ms
faster with NFC (P ≈ 0.001)
► On average, subjects older than 30 are faster with NFC
NFC vs Touchscreen Based Interaction
Evaluation
43/38
44. Task efficiency (IV)
►
Subjects who own a touchscreen-phone are clearly
faster with this kind of interaction
► 18
►
seconds faster on average (P < 0.001)
While it is not that clear that subjects with previous
experience with a NFC phone are faster
►6
seconds faster on average (P ≈ 0.05)
NFC vs Touchscreen Based Interaction
Evaluation
44/38
46. Conclusion (I)
►
Platform to access the most-used services in mobile
phones using NFC.
►
New approach for NFC-based interaction: Actions as
combinations of tags.
►
Proposed interaction scheme considered convenient and
useful by test users.
NFC vs Touchscreen Based Interaction
Conclusion
46/38
47. Conclusion (II)
►
NFC also proved faster than touchscreen, specially for…
► Older
people (over 30 years old)
► Those less familiar with touch-based interaction
► Those less familiar with new technologies in general
►
Learning curve of NFC-based interaction also proved to
be low.
NFC vs Touchscreen Based Interaction
Conclusion
47/38
48. Future Work
►
Run a larger scale experiment
► Larger
number of subjects, more homogeneous groups
► More
subjects in older age groups
► More people with lower technological skills
► Minimize
►
dependency among variables
Longer in time
► Study
the long-term learning effect in both interaction
approaches
►
Apart from studying task efficiency, measure interaction
errors with both technologies
NFC vs Touchscreen Based Interaction
Conclusion
48/38
50. All rights of images are reserved by the original
owners*, the rest of the content is licensed under a
Creative Commons by-sa 3.0 license.
*
Android Open Source Project: http://en.wikipedia.org/wiki/File:Nexus_4.png
http://en.wikipedia.org/wiki/File:Nexus_10.png
Stefan Svartling: https://picasaweb.google.com/lh/photo/LJ44WQRTFlHX5ciB_xLV8lDUMCdxqHK-2KnxRTTdAcc
Google Inc.: http://commons.wikimedia.org/wiki/File:Google_Chrome_icon_(2011).svg
Photoshopedia: https://www.iconfinder.com/icons/33896/camera_photography_webdesign_icon
Wwalczyszyn: https://www.iconfinder.com/icons/67500/android_maps_r_icon
Aha-Soft:
https://www.iconfinder.com/icons/54522/feed_garbage_good_tidings_journal_lection_literary_garbage_mandarin_mandarine_news_newspaper_or
ange_organ_paper_print_read_reader_reading_retiree_rss_sheet_slipslop_tangerine_tidings_uncos_yellow_icon
https://www.iconfinder.com/icons/54526/community_connection_consultation_consulting_earth_global_group_internet_large_group_network_polar
_round_table_social_social_network_users_world_icon
Alessandro Rei: http://findicons.com/icon/254687/audacity
Oliver Scholtz (and others): https://www.iconfinder.com/icons/24246/and_calendar_preferences_tasks_icon
Bharathp666: https://www.iconfinder.com/icons/72149/android_base_gmail_icon
Juan José Aza: http://www.flickr.com/photos/todojuanjo/2630161117/
Chris Arvin: http://dribbble.com/shots/396641-Freebie-PSD-Android-4-0-UI?list=users
W3C: http://www.w3.org/RDF/icons/rdf_flyer.svg
TAPPIN: http://tappinn.com/public/images/main2011/nfc-sticker.jpg
Icons Land: https://www.iconfinder.com/icons/73049/base_chartreuse_inside_map_marker_socialmediabookmark_icon
Everaldo Coelho: https://www.iconfinder.com/icons/18095/clouds_sun_weather_icon
NFC vs Touchscreen Based Interaction
50/38
51. Mobile NFC vs Touchscreen Based Interaction:
Architecture Proposal and Evaluation
Koldo Zabaleta
{koldo.zabaleta@deusto.es}
DeustoTech - Deusto Institute of Technology, University of Deusto
http://www.morelab.deusto.es
NFC vs Touchscreen Based Interaction
51/38