Employing Gamified Crowdsourced Close-Range Sensing in the Pursuit of a Digital Twin of the Earth

Figures & data

Figure 1. A screenshot of the spider vacuuming application.

Figure 2. A screenshot of the spray painting application.

Figure 3. A screenshot of the point cloud visualization application.

Table 2. Interview guide.

Theme	Questions
General	Could you please describe each of the four scenarios you just went through?
Preference	Which of the four scenarios did you prefer the most and why?
	Which of the four scenarios did you prefer the least and why?
Scenarios	Did you understand what you were doing in the background during the scenarios?
	How would you improve on each of the scenarios?
Conclusion	Is there anything you would like to add to what was already said?

Table 1. Data sources collected for this study.

Source	Description
Point cloud and mesh data	For all participants, we stored the point cloud and mesh data that they generated in each of the four experiments separately.
Video recording of the activity	We used the screen capture functionality of the experiment devices to record the screen during the experiments.
Voice recording of the activity	We asked participants to think out loud as they were using the four applications, and we recorded their thoughts using the voice memos application on the experiment devices and later transcribed them.

Table 3. Demographic characteristics of the participants and the environment for their experiment.

Participant	Age	Gender	Education level	Experiment environment
P1	30–39	Male	Master’s degree	Forest in fall
P2	20–29	Female	Bachelor’s degree	Forest in fall
P3	40–49	Male	Doctoral degree	Forested landscaping in fall
P4	30–39	Female	Doctoral degree	Forested landscaping in fall
P5	50–59	Female	Doctoral degree	Forested landscaping in fall
P6	50–59	Male	Doctoral degree	Forest in fall
P7	60–69	Male	Doctoral degree	Forest in fall
P8	40–49	Male	Doctoral degree	Forest in winter
P9	30–39	Female	Master’s degree	Forest in winter
P10	30–39	Female	Master’s degree	Forest in winter
P11	30–39	Male	Doctoral degree	Forest in winter
P12	20–29	Female	Master’s degree	Forest in winter
P13	30–39	Female	Master’s degree	Forest in winter
P14	30–39	Female	Master’s degree	Forest in winter
P15	50–59	Male	High school diploma	Forest in winter
P16	40–49	Female	Master’s degree	Forest in winter

Figure 4. Ground points and non-ground points of P12.

Figure 5. Most and least favorite AR scenario.

Table 4. Summarizing a crude estimate of the effects of the four gamification designs on three key categories in terms of point cloud generation: movement, focus on detail, and vertical change.

Scenario	Movement	Focus on detail	Vertical change
Nothing	Low	Low	Low
Spider vacuuming	High	Low	Medium
Spray painting	Varied	Varied	Varied
Environment scanning	Low	High	High

Figure 6. A Boxplot showing the variation in the point cloud-derived metrics by the applied gamified scenario. On each box, the red vertical line indicates the median and the bottom and top edges of the box indicate the 25th and 75th percentiles, respectively. The whiskers extend to the most extreme data points not considered outliers, and the outliers are plotted individually using the “+” marker symbol.

Table 5. The p-values of paired-sample t-tests quantifying statistical differences in the point cloud metrics between gamified scenarios. Bolded values imply statistical significance in the observed difference at 95% confidence (i.e., p < 0.05).

Comparison	No. points	Ground points % / non-ground points %	XY projection area (m^2)	Z range (m)
#1 vs. #2	0.027	0.014	0.070	0.066
#1 vs. #3	0.001	0.141	0.049	0.406
#1 vs. #4	0.001	0.584	0.086	0.225
#2 vs. #3	0.007	0.006	0.536	0.315
#2 vs. #4	0.002	0.026	0.292	0.003
#3 vs. #4	0.118	0.427	0.528	0.045

1 = Nothing, 2 = spider vacuuming, 3 = spray painting, 4 = environment scanning.

Table 6. Suitable interactions for specific goals.

	Low density capture	High density capture
Focus on ground	Spider vacuuming	–
Focus on above ground	Nothing	Spray painting, environment scanning