Development of a platform using Machine Learning to determine the optimal harvest time window for hemp fields

donderdag 12 januari 2023
Hemp field Hemp field
Developed with ASP.NET Core Developed with ASP.NET Core
Web frontend developed with Angular Web frontend developed with Angular

The Challenge

Climate change and the Green Deal requirements will be a catalyst to work more with natural fibre crops in agriculture, like hemp. In this project the challenge is help the farmers dealing with hemp to optimise their yield.

The goal

The goal of the HarvestNow platform is to help the huge potential of flax to be fully used, by helping the farmers in determining the best time window to harvest a specific field of hemp. It is a very easy to use application which will be used by flax farmers to determine the optimal time to harvest on their fields. The tool combines several external sources (eg ground sensors, weather forecasts,...) to calculate the optimal time window. The platform is focused on flax fields, because expert knowledge of the flax farmers is integrated in the application. It aims to achieve the best possible yield for the farmer and the best possible product for further processing.

Machine Learning

To determine the optimal harvesting time window, the historical measurements (soil, temperature, ...) must be used to generate a forecast. To this end, a Machine Learning model is used based on the SSA (Singular Spectrum Analysis) algorithm. SSA is a powerful framework for decomposing the time-series into trend, seasonality and noise components as well as forecasting the future values of the time-series. In principle, SSA performs spectral analysis on the input time-series where each component in the spectrum corresponds to a trend, seasonal or noise component in the time-series. It generates a trendline into the future that allows the application to pinpoint the harvesting time.



Teillage Dewynter logo

Teillage Dewynter logo

Demeter logo

Demeter logo

The HarvestNow platform was developed by TRI-S on behalf of the project leads, ODYC and SAS Dewynter. The platform is developed as a web frontend where the fields can be defined using a GeoJSON upload and extra data can be inputted. The harvest time can be visualized in this frontend or the field data, including the optimal harvest time window can be requested via an API.

The API service is registered in Demeters' Brokerage Service Environment and the Demeter Enabler Hub, making it available to third parties who'd wish to use the data.


DEMETER’s goal is to lead the digital transformation of Europe’s agri-food sector through the rapid adoption of advanced IoT technologies, data science and smart farming, ensuring its long-term viability and sustainability. 60 partners from 18 European countries are involved in the DEMETER project.

The main website can be found here.

DEMETER is a Horizon 2020 project (857202) supported by the European Union.