Cognitive services in agriculture
Cognitive services and the application of AI-based methods are a central component of the Fraunhofer lighthouse project “Cognitive Agriculture”. As we understand it, a cognitive system is first and foremost a system that behaves intelligently – from a human perspective – in a situation, for example by adapting its behavior depending on the context.
This can, but does not need to, include neural networks or Machine Learning methods as technology. It can also simply mean that several data sources are used and combined to make the triggering of an action safer, more sustainable, and more transparent – or simply that one or several context factors are included in the consideration, such as available sensor values.
In this sense, the key to digital transformation in agriculture lies in the availability of machine-readable data and information. These should be available in high quality, and there must be an understanding of the interrelationships. Only on this basis can cognitive services be developed, that is, systems that offer an interface that is helpful and feels natural to humans to support them in their daily work.
Agricultural solutions with cognitive services
In our lighthouse project, cognitive services are linked to an agricultural solution or a concrete service that transforms data into information and decision support.
Among others, we are working on solutions for the following topics:
- Representation and control of the nitrogen balance in the nutrient cycle
- Support for robust harvest campaign planning
- Optimization of the use of fertilizers through predictive plant cultivation
Depending on the state of the implementation, results on the cognitive solutions are presented here in more detail.