Use Case

Blooming business

The challenge

Understanding the current state of your crops is vital to producing optimal crop yields. In a world that keeps getting hotter and drier, you need to identify problems as soon as they arise. The agricultural sector is under high pressure to increase crop yield and to maximize the use of farming land that is available. One of the consequences is that greenhouses grow larger and fuller. In many cases, one farmer is responsible for thousands of square meters with a diversity of crops making it very hard, costly, and complex to inspect regularly. In the past one relied on gut feeling and experience. But with the current changes in climate and increased regulations more data driven inspections are needed. To solve this problem our team and Applied Drone Innovations (ADI) joined forces to develop an autonomous inspection system to efficiently monitor plants in greenhouses. 

The solution

The autonomous monitoring system consists of ourVertex drone platform equipped with a camera and artificial intelligence from ADI. To facilitate navigation inside the greenhouse, we used our proprietary Starling navigation system, which allows the drone to fly autonomously and to collect image data of the plants. These images are analyzed for crop growth stages and plant diseases using the artificial intelligence from ADI.

Flying inside greenhouses is challenging. First because (RTK-)GNSS positioning is not possible due the weak signal strength inside the greenhouses. This requires an alternative positioning system than that what is typically used in outdoor crop inspection use-cases. Given the cluttered environment inside the greenhouses an accurate and robust navigation and perception system is needed to avoid collisions. Also, the warm temperature and high humidity in these types of spaces pose another challenge that needs to be overcome. Any drone and the navigation infrastructure need to adapt for long term deployment under these conditions.

In turn, ADI developed the AI to analyze visual data registered by training the AI to recognize healthy and unhealthy plants. This is done by feeding the AI a large quantity of training pictures of healthy plants. But also, plants that are infected with the most common 5 crop diseases. 

More benefits of using drones for crop inspection:

  • Save time
  • More regular inspections
  • Provide better visual data