Research project New technologies and artificial intelligence for optimal crop management in floriculture

In progress SIERTECH 01/10/2024 → 30/09/2028

Contact our expert

Peter Lootens

Precision farming and phenotyping expert

Contact

Info Contact Partners

General introduction

This project targets more automated crop monitoring methods in the floriculture sector in Flanders, using the most appropriate cameras, drones and AI (artificial intelligence) applications. The floriculture sector in Flanders is highly productive and operates in a highly competitive market. Ornamental growers must constantly try to make the right management decisions during cultivation: growth, development, stress, diseases and pests, effectiveness of pest treatments, quality and and also availability of plants for sale. All this monitoring is mostly done based on the expertise/experience of one or a limited number of people in the company. The continuous trend towards scaling up, the diversity of crops within one company, and the insufficient availability of personnel with the necessary expertise make it more challenging to closely monitor all plant batches within a company.
The goal of SIERTECH is therefore to develop crop monitoring methods based on the intelligent automatic interpretation of multifaceted camera images, and this both for crops under glass and in open air.

Research approach

We collect data with different types of cameras (visual, multispectral, thermal). We convert these into relevant information to support management decisions during cultivation. We work in three areas: (1) crop monitoring (growth, development, stress, ... possibly linked to a growth model); (2) monitoring of diseases and pests and the effectiveness of pest treatments, (3) inventory and monitoring of quality and availability of plants for sale. We are always developing tools that are able to say what action should happen in what place at what time for species-specific advice.

Relevance/Valorization

Embracing AI tools and applying them in floriculture operations will only become feasible if the research community helps develop and train the systems. The industry is experiencing a growing need to have smart farming technologies available. The likelihood of implementation is high, as is the expected social added value: All developed tools aim at a reduction in labor costs by automating certain actions, better stock monitoring and consequently a better grower-customer relationship, a reduction in resource use through early disease and pest detection, reduced waste and better quality.