PROJECT | IOF 2020
Arable farming represents the largest agricultural sector in the EU in terms of acreage (60% in 2013) and number of primary production holdings. The sector has important ambitions, such as increasing production for food, feed, bio-based products and energy with the same or less input. At the same time, the sector faces several challenges. Stopping loss of soil fertility, prevent the pollution of groundwater and tackle disease/weed resistance for example. Against this background, a more sustainable food value chain is a necessity.
IoT technology enables precision farming. In the use cases we link existing sensor networks, earth observations systems, crop growth models and yield gap analysis tools to a variety of databases. This combination of information creates effective, standardized actuation protocols (‘task maps’) for machines and robots. Focusing on the cultivation of three main crops (wheat, soybeans and potatoes), in different European regions and climate zones, the trial includes activities along the cropping cycle. With the help of IoT technologies, data on key variables such as the soil, climate conditions, growth of plants and weed, disease or pest prevalence can be combined in a meaningful way.
The innovative farm management systems in the arable farming trial address urgent challenges, such as the efficient use of pesticides, fertilizers and energy. The smart combination of data also enhances transparency and food safety along the food chain.
In the framework of the IOF2020 European project, ARVALIS and Orange are in charge of a use case dedicated to wheat crop management. Experiments have been set in the Beauce region and in the South East to demonstrate the feasibility of such service and evaluate the added value in comparison with more classical practices. The CHN crop model (ARVALIS) has been chosen because it simulates with good accuracy both the crop development and the nitrogen budget in the soil. Correction measurements are regularly acquired using the SENTINEL2 satellite images (ESA) and Field Sensor (BOSCH), an IoT system set in the field that daily measures the crop growth. A fusion procedure between both data sources has been developed (HIPHEN and INRA) to provide high spatio-temporal resolution observations. Measurements are then assimilated in the CHN model using a Kalman filter. Last, decisions of timing and quantity of nitrogen application is taken by analyzing the simulated and forecasted crop NNI (nitrogen nutrition index).
The first results obtained in 2018 show that this approach increases the efficiency of the nitrogen applications compared with classical management. Results using the corrected model are over-performing the use of the model without any in season sensor measurement. In 2019, a larger scale experiment allows a more robust evaluation of economic and environmental gain. In addition, this dynamic approach allows accounting for other constraints like weather conditions or work organization. It is possible to test several strategies and assess the impact of differing or anticipating a nitrogen application, that was difficult to do with more classical management that rely on specific development stages.