Pcfruit, VITO and FlandersMake launch EIP Operational Group project for yield prediction by flower counting in strawberry plants
Transforming strawberry farming with drones, sensors and AI
In a groundbreaking effort to enhance strawberry cultivation, pcfruit, in collaboration with VITO and FlandersMake, has launched an innovative 2-year EIP Operational Group project focusing on yield prediction based on flower counting in everbearing strawberry growing in substrate under protective tunnels. This initiative leverages advanced technologies, including mini drones, moving sensors, and AI image analysis, to address the inefficiencies and inaccuracies of traditional yield prediction methods. Mini drones, equipped with high-resolution cameras, navigate through the strawberry tunnels capturing detailed images, while moving sensors gather real-time data on flower density. These data points are then processed by AI algorithms to accurately identify and count flowers, which are one of the input data for a yield prediction model. This approach not only reduces the time and labor involved but also ensures precise and consistent flower counts, leading to more accurate yield predictions, better harvest planning (and harvesters' planning), as well as improved marketing (auction prices).
3-step approach
A previous collaborative project has already demonstrated promising results in everbearing strawberry growing in full soil in open air, showcasing increased accuracy in flower counting and improved yield predictions for participating farmers. The AI model for flower detection in open air strawberry cultivation is integrated and available on the MAPEO drone processing platform of VITO.
As this growing system is rapidly being replaced by a system in substrate on scaffolds in plastic tunnels, this new project aims to set a new standard for precision agriculture by integrating these cutting-edge technologies and making strawberry farming more efficient. To achieve its objectives, the project will implement a comprehensive three-step approach.
-
First, mini drones and moving sensors gather high-resolution images from the strawberry tunnels.
-
Next, AI algorithms meticulously analyze these images to identify and count flowers which are used together with other input data to generate predictive yield data.
-
Finally, this processed data delivers actionable insights, providing farmers with precise information about flower density and expected yields 3 to 4 weeks later, enabling them to make informed decisions in advance for optimal crop management including harvest organisation and marketing in collaboration with the producer organisation.
Key benefits
The project is expected to significantly enhance efficiency, accuracy and cost-effectiveness, making strawberry farming more productive and in the end environmentally friendly.
Here are the key benefits in detail:
- Efficiency: Significantly reduces the time and labor required for large scale and thus, reliable flower counting
- Accuracy: Ensures precise and consistent flower counts, improving yield predictions and human resource management.
- Cost-Effective: Lowers labor costs and minimizes the need for manual intervention. In addition, more accurate yield prediction allows for better marketing potentially increasing the income for the farmer.
- Sustainability: With these technologies we can also optimize resource use, contributing to more sustainable farming practices in the near future.
