IEEE Robotics & Automation Magazine - December 2015 - 93

(a)

(b)

(c)

Figure 4. An example of typical user interaction data over the application screen for (a) manual, (b) semiautonomous, and
(c) autonomous.

User Response (%)

use the iDevice application,
he/she could have been moti25
21.4
What Issues Are You Concerned About?
vated to control the fish, and
20
What Issues Do You Want to Learn More About?
decided to use one of the control modes rather than let the
15
12.9
robot passively swim in the
11.2
10.7
9.4
autonomous mode.
10
7.1
6.7
6.7
5.8
5.4 5.4
To examine the signifi5
cance of mean time usage of
1.3
the control modes, a one-way
0
Marine
Climate
Marine
Marine Overfishing Ocean
analysis of variance (ANOVA)
Acidification Habitat
Change
Animal
Pollution
test was used [21]. The
Destruction
Conservation
ANOVA test rejected the null
hypothesis that the variances Figure 5. The percentage of user responses to the degree of concern for various environmental
in time usage between the issues (N = 121) and degree of user curiosity to learn about environmental issues (N = 112). N
control modes were equal indicates the number of users who responded to the respective question.
^ F2, 669 = 185, p 1 0.01 h [21].
Futhermore, a post-hoc Tukey's honest significant difference
Table 1. The mean and standard deviation of
(HSD) test indicated that there were significant differences
the age of visitors who answered the survey
^
h
between each pair of control modes p 1 0.05 . One explaquestions and the breakdown for selected
nation for these results may be that visitors at informal setenvironmental issues.
tings are more engaged in hands-on approaches [22], [23],
Environmental
Responses to
Responses to
which was available to them in the manual and semiautonoIssues
Concern on
Learning More
mous mode, as opposed to the passive approach in the autonIssues
About Issues
omous mode.
Mean Standard Mean Standard
Figure 4 shows a the usage of a typical visitor using the
Deviation
Deviation
iDevice and touch locations on the screen. The touches are
All choices
13.9 10.7
13.9 10.9
overlaid on the iDevice screen, indicating that in the manual
Marine pollution
12.8 10.4
12.5
9.7
mode a user tended to utilize the steering and speed control
Overfishing
15.6 10.9
15.1 11.7
sliders. In the semiautonomous mode, the touch interactions
Ocean acidification
8.7
3.1
12.1
8.6
were more widely varied across the screen, as creating a traMarine
habitat
13.5
10.2
8.4
4.8
jectory required only a few taps. Not surprisingly, in the
destruction
autonomous mode, users' touches were sparse and concenClimate change
17.1 13.6
18.6 13.8
trated on the animation, as the robotic fish did not react to
Marine
animal
11.3
7.4
13.2
9.7
users' touches. The interactions with the control modes of the
conservation
application suggest that the user was able to navigate through
the application and explore the different control modes.
Data on the environmental issues of concern were taken that users were most concerned about were marine pollution
from those users who completed this section of the survey, and (21.4%), overfishing (10.7%), and climate change (9.4%) (Figpercentages were calculated out of the total number of users. ure 5). In addition, the top three environmental issues that users
According to users' reports, the top three environmental issues would like to learn more about were marine pollution (12.9%),
DECEMBER 2015

IEEE ROBOTICS & AUTOMATION MAGAZINE

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - December 2015