Successfully growing high-value crops such as fruit (citrus, apples, grapes, cherries, kiwis, table grapes, pomegranates, and more) and nuts (almonds, pecans, pistachios, and more) requires farmers’ continuous attention. If a problem is not detected, the crop yield could be significantly reduced. Sensors provide the foundation and play an essential role in the evolving detection processes.
The first step: sensing and drones
For a more complete crop analysis, sensors must see both the visible and the invisible. To do this, some systems sense two forms of sight. Visible images come from a red-green-blue (RGB) camera with a multispectral camera and even thermal infrared (IR) imaging added to scan and analyze crop growth. One supplier offers a 20 MP RGB image camera. With the sensors installed on an aerial drone, crops can be viewed in a manner that allows more thorough analysis than ever before so farmers can monitor, manage, and protect them.
For example, the trees in an orchard can be viewed from above, and any tree or block of trees with signs of lower health can be treated as soon as possible. With frequent, ongoing observations, especially after treatment, multiple drone flights can reveal the success of the treatment. For large orchards, drone flights can measure hundreds of acres, making crop analysis a wise investment for a successful season. The drones are operated either by the grower or through a third-party pilot network.
The next step: add AI
In addition to maximizing crop growth and yield, monitoring the multispectral and thermal imagery can help reduce the amount of water and ensure that it goes to the crops most efficiently and uniformly. The same technique can be used for fertilization.
To do this, data analysis through hyper-localized forecasting models allows an artificial intelligence (AI) system to protect the data forward toward harvest time. With accumulated data, the models for a specific farm are fine-tuned to its specific environment.
With AI models learning and adapting to localized growing conditions, forecasting accuracy improves, and comparisons to previous years are possible. For even deeper analysis, a separate AI program provides a digital model of each tree on the farm and tracks it over time. By considering each tree as a machine in a factory, its production can be optimized to provide the highest-quality fruit.
Add another detail
With the data already available from the sensors on drones and stored in the cloud, AI can also provide another key diagnostic and remediation: identify and prevent bug problems. By accurately identifying the invading pest(s) and targeting the response with AI, a uniform dispersal of various beneficial insects from the drone can accurately attack the pests on a per-tree level with a 50m search radius and 20m buffer in one system. Another AI provider stated that their AI system has reduced the monitoring required for every tree for pests and diseases on a 50-hectare farm from an entire day to just 20 minutes.
Down-to-earth sensing
Farmers have been using state-of-the-art soil sensors for many years. The large amount of raw data provided by these ground-level sensors can be added to the new input from above and give an even more complete interpretation and real-time AI analysis.
References
Aerobotics
Precise beneficial insect applications=
DJI Mavic 3M
CropX Case Study (amazon.com)
Leave a Reply
You must be logged in to post a comment.