Geofencing with R for Enhanced Data Collection in Kobotoolbox

Geofencing is a powerful tool that combines location technology with data analysis allowing you to automate and check data collection within predefined geographical boundaries.

R
Kobotoolbox
Data Collection
Author

Oluwatosin Orenaike

Published

January 5, 2024

Introduction

Geofencing is a powerful tool that combines location technology with data analysis, allows you to automate data collection within predefined geographical boundaries. I will show you a step-by-step guide, using R, to integrate geofencing with Kobotoolbox and boost your data collection efficiency. This process involves two major steps:

  1. Getting the vertex for your location of interest.
  2. Uploading vertex to KoboToolbox (link to a video).

Imagine this:

  • You enter a designated study area, and data collection triggers automatically. No more remembering to press buttons or check locations!
  • Your data comes from precisely where you need it, thanks to the power of geofencing boundaries. No more worrying about stray data points or missed locations!
  • Kobotoolbox forms adapt based on location, with possibilities of dynamically changing questions or displaying relevant information for specific areas.

Let’s dive in!

Load data

Step 1: Shape Up Your Boundaries

First, define your geofencing areas. Grab the shapefiles for your target zones — these handy files store information about geographical boundaries. Our example uses a sample shapefile named “Sample_locations” with details like Sites, Zones, House, and Blocks. Download it here: [link to your sample shapefile]

# Load your shapefile (ensure you provide the correct path)
Sample_locations <- st_read("./data/shp_file/Sample_locations.shp")
Reading layer `Sample_locations' from data source 
  `C:\Users\orenaike\OneDrive\01_ORENAIKE\02_CAREER_AND_DEVELOPMENTS\01_Schools\Web_Development\Portfolio\otomisin.github.io\posts\geofencing\data\shp_file\Sample_locations.shp' 
  using driver `ESRI Shapefile'
Simple feature collection with 6 features and 4 fields
Geometry type: POLYGON
Dimension:     XY
Bounding box:  xmin: 605820.6 ymin: 853074.4 xmax: 606177 ymax: 853555.9
Projected CRS: WGS 84 / UTM zone 35N
# Check and plot the data table to understand its structure
print(st_drop_geometry(Sample_locations))
  OBJECTID Zone_ Block   House
1        1     B     1 House A
2        2     B     2 House B
3        3     A     1 House C
4        4     A     2 House D
5        5     A     3 House E
6        6     A     4 House F
# Plot the locations
Sample_locations |>
  ggplot() +
  geom_sf() +
  geom_label(aes(x = st_coordinates(st_centroid(geometry))[, 1], 
                 y = st_coordinates(st_centroid(geometry))[, 2], 
                 label = House), 
             size = 3, fill = "lightblue", color = "black") +
  theme_void()

.

Step 2: Extract Those Coordinates

Now, use R’s st_geometry function to extract the precise latitude and longitude values from your shapefile. We’ll store these coordinates in a separate data frame for easier manipulation.

# Extract coordinates using the same Coordinate Reference System (CRS) as the original data
polygon_vertices <- lapply(st_geometry(Sample_locations), st_coordinates)

# Map each polygon_vertices list to its OBJECTID row
mapped_data_v1 <- Map(function(vertices, objectid) {
  data.frame(OBJECTID = objectid,
             p_longitude = vertices[, "X"],
             p_latitude = vertices[, "Y"])
}, polygon_vertices, Sample_locations$OBJECTID)

# Combine the mapped data frames into a single data frame
mapped_data_v1 <- do.call(rbind, mapped_data_v1)

# Integrate the mapped data with the original shapefile data
Sample_locations_points <- Sample_locations |>
  as.data.frame() |>
  left_join(mapped_data_v1, by = "OBJECTID") |>
  st_as_sf(coords = c("p_longitude", "p_latitude"), crs = st_crs(Sample_locations))


summary(Sample_locations_points)
    OBJECTID        Zone_              Block              House          
 Min.   :1.000   Length:66          Length:66          Length:66         
 1st Qu.:1.000   Class :character   Class :character   Class :character  
 Median :2.000   Mode  :character   Mode  :character   Mode  :character  
 Mean   :2.667                                                           
 3rd Qu.:4.000                                                           
 Max.   :6.000                                                           
          geometry 
 POINT        :66  
 epsg:32635   : 0  
 +proj=utm ...: 0

Step 3: Generate Your ID Nodes

Sample_locations_points_v1 <- Sample_locations_points %>%
  group_by(Block) %>%
  mutate(id_node = paste0(Block, "_", row_number())) %>%
  ungroup()

# View the points and polygons on a map
ggplot() +
  geom_sf(data = Sample_locations) +
  geom_label(data = Sample_locations_points_v1,
             aes(x = st_coordinates(st_centroid(geometry))[, 1],
                 y = st_coordinates(st_centroid(geometry))[, 2],
                 label = House),
             size = 3, fill = "lightblue", color = "black") +
  geom_sf(data = Sample_locations_points_v1) +
  theme_void()

Step 5: Export the Spatial Data for Kobotoolbox Integration

write.csv(Sample_locations_points_v1, "./Sample_locations_points_v1.csv")

# The exported CSV file can now be used in Kobotoolbox for enhanced data collection with geofencing

Step 5: Integrate with Kobotoolbox

Congratulations! You now have a spatial data frame enriched with precise coordinates and unique identifiers, ready to be seamlessly integrated with Kobotoolbox. Unleash the power of location-aware data collection, with automatic form triggers and data collection tailored to specific geographical zones.

Beyond the Code: Here are some additional tips to take your geofencing journey to the next level:

  • Real-world examples: Think about using geofencing to study air quality in specific city districts, automatically triggering data collection at designated times.
  • Challenges and solutions: Consider potential challenges like battery drain on data collectors and GPS limitations. Optimize data collection forms and schedule strategically to mitigate these.
  • Dive deeper: Explore our GitHub repository (link here) for the full code buffet and detailed tutorials to become a geofencing master.

Don’t let your data collection be stuck in the manual age! Leverage the magic of geofencing with R and Kobotoolbox to supercharge your fieldwork efficiency and precision and experience the power of automation, precision, and streamlined workflows!

You can check how to integrate the output into Kobotoolbox. Check this video on how to.

Conclusion

Don’t let your data collection be stuck in the manual age! Leverage the magic of geofencing with R and Kobotoolbox to supercharge your fieldwork efficiency and precision, experiencing the power of automation, precision, and streamlined workflows.

For detailed integration with Kobotoolbox, check this video on how to.