Blog

Notes from the “Drones for Agriculture” Knowledge Sharing Day

As a side event of the 2018 FOSS4G Conference, CGIAR-CSI, WeRobotics, and Tanzania Flying Labs co-organized a "Drones for Agriculture" Knowledge Sharing Day event at the Tanzania Data Lab on August 23, 2018.

As a side event of the 2018 FOSS4G (Free and Open Source Software for Geospatial) Conference, our CGIAR-CSI community co-organized a “Drones for Agriculture” Knowledge Sharing Day on August 23, together with WeRobotics and Tanzania Flying Labs, at the Tanzania Data Lab (dLab), based in the University of Dar es Salaam Computing Centre.

With the facilitation of Sonja Betschart (WeRobotics), about twenty avid drone users were presented with the demos of hardware (quadcopter, fixed wing; different sensors)  software (Pix4Dfields), followed by a series of  knowledge sharing sessions where everyone shared diverse use-cases in agriculture and challenges, opportunities, and how we, as a community, can make the best use of drones in agricultural research and development.

Use-Cases & Opportunities

Each participant presented how they are using drones and answered many questions others asked on how each use-case can successfully be replicated in other areas. Presented use-cases included:

  • Use drones to acquire high-resolution imagery in-season to calibrate satellite imagery and estimate pre-harvest crop losses
  • Drone-captured imagery to help farmers to monitor crop growth and optimize farm management decisions (e.g., fertilizer and planting)
  • High-resolution mapping to identify suitable areas for planting in marginal environments
  • Ecological monitoring and detecting changes in the environment (e.g., encroachment, fire, and logging)
  • Anti-poaching
  • Monitor human-wildlife conflicts and wildlife damage on agriculture
  • Verifying crop insurance claims
  • Real-time mapping of crops within mixed land-use areas (e.g., bananas)
  • Monitoring of crop diseases

Through the free-flow discussion following presentations, two important potential use-cases emerged. One was on the use of drone imagery to support crop breeding scientists to measure key traits of field crops, such as crop height, canopy size, growth stages, over large areas effectively. The second one was the automated mapping of farm plot boundaries and sizes. Both use-cases were discussed as the high priority in agricultural research and considered readily feasible in the very near future. Participants started brainstorming on these two topics right away.

Challenges & Solutions

During this session, participants discussed what challenges they faced in the use-cases and discussed what solutions there might be to overcome at various levels. Key challenges identified included many issues that can potentially be overcome with technical training, such as:

  • Not enough training opportunities on, for example, drone operations, best practices, troubleshooting, and imagery analyses
  • Not easy to engage with local communities in the rural areas
  • Not easy to manage and share (huge) drone imagery data across
  • Difficult to cross-calibrate drone imagery with satellite-captured remote sensing data
  • Difficult to keep the consistency in the drone imagery across time and sensors

Additionally, some practical challenges were identified as something we, the drones in the agriculture community, can possibly influence (albeit indirectly), including:

  • Handling of potentially sensitive data with privacy issues
  • Importation of drone equipment from abroad
  • Securing permits and complying regulations
  • Cost of equipment insurance

For most of the challenges (other than something we can’t handle — such as weather conditions), the discussion amongst participants converged to the realization that we, as the practitioners in agriculture, are not in the best position to handle these complex issues. Technology and the regulatory issues, along with data management and privacy concerns, are all dynamically evolving. It is important for the community to understand where the issues are and what options exist. However, the most sensible and logical approach would be the establishment of a technical partnership with a drone professional, Flying Labs — such that we can focus on how best we can analyze the drone-captured data and generate useful information for making impacts on the ground.

Code of Conduct

CodeOfConduct2

The full-day program concluded with the participants’ pledge to the Drone Code of Conduct for Social Good, developed by WeRobotics based on the Humanitarian UAV Code of Conduct. Above anything, we all agreed to fly drones responsibly, be prepared, ensure not to harm our beneficiaries and environment, and be accountable. Just like Uncle Ben once famously said, we were all reminded that “With great power comes great responsibility!”