PrecisionFarming

Important results from the project

The project´s aim was to develop an AI system that can detect improvements in our cultivation process that in turn can contribute to improvements in energy efficiency, as well as higher quality and quantity of the products we grow. At the end of the project there is a rudimentary, proof-of-concept version of such an AI system where validation still remains. The project has also generated an instructional guide for future data storage and handling that takes into account the large amount of data which will accumulate.

Expected long term effects

The project has resulted in a proof-of-concept for specialized data collection as well as a basic AI system that, using ML algorithms, transforms images of our products into quality assessments. We expect this basic AI system, with further development, can match quality assessment with data from input parameters in our cultivation, and give suggestions for improvements.

Approach and implementation

We intended to utilize Grönska´s collected data on input parameters in our cultivation and using the supervised ML method, develop an AI that analyzes and evaluates input parameters and outcome data. Early in the project, we discovered supplementary needs in the form of outcome data. Proof of concept on solution for collecting this data therefore became an early focus. This work has partially delayed results of the project or rather led to results on a smaller scale than planned. However we see huge opportunities to continue our AI-journey with the results generated from this project.