Increasing demand for food puts pressure on agricultural production. Fertilizer use is one of the essential factors for increasing agricultural yield. Global demand for fertilizer is expected to reach 200 million metric tons per year, representing a ten-fold increase since 1950. This will accelerate negative impacts on the environment and affect long-term health of agricultural fields. Therefore, fertilizer usage must become increasingly efficient.
Precise fertilization requires a deep understanding of issues with so-called plant nutrient deficiency (PND). Monitoring PND and subsequently providing information in real-time, fertilization requirements is our main objective in developing this IoT-based, integrated service.
Fertilization efficiency highly depends on PND diagnosis and Field Nutrient Budgeting (FNB). There are several well-known methods and tools available in the market for these issues. However, none of these offer precise PND diagnosis, variable rate fertilization optimization functionality and satellite-based fertilization map creation at once. This is a very complex process for farmers as they must collect these crucial information with different tools.
To address agricultural productivity under the frame of an environmentally friendly concept, SIITAg provides an intelligent system that can identify PND and generate a variable-rate fertilization application map for the optimal agricultural output. The system is cost-effective, scalable and globally available for farmers.
Users and their needs
SIITAg addresses three groups of potential users with their specific needs, which will be in our focus.
- Strict restrictions on fertilizer usage and the need for cost-effective production create pressure for farmers to optimize their fertilizer usage.
- Farmers need user-friendly access to such information without having to purchase complex and costly machinery to get it.
Additionally, two groups of intermediary customers have their own specific needs that are met by SIITAg.
- Agricultural consultants
- Consulting farmers on their fertilization demands increasingly specific, individually-tailored advice. This requires technology that can identify field management zones on a large scale, using accurate PND information.
- Agribusinesses (agrochemical and -machinery companies)
- Strong customer relationships are essential for a scaling business. To provide their customers with valuable additional services and stand out from the competition, companies require state-of-the-art technology and knowledge coupled with easy-to-use interfaces.
- To effectively design their supply-chain and marketing strategies, companies need to monitor how the demand develops among their customers. Rich, meaningful data from large numbers of users can provide those insights.
SIITAg service is designed to be available worldwide. Internally the service will be piloted in Germany. Later, the service will be available for the EU countries, North America and South America consecutively.
Service/ system concept
Using the SIITAg service, farmers can visualize their fertilizer needs using fertilization recommendation maps and manipulate these recommendation maps according to their knowledge of the spatial variability of their fields. The web portal allows farmers to get insights about their field performance and enhance their agricultural decision-making processes significantly.
The SIITAg front-end will operate as a smartphone app. The backend system service design employs artificial intelligence and a machine learning algorithm to combine satellite data with data collected through the app to generate the recommendations. The process is as follows: farmers will use the smartphone app to take pictures of a few crop leaves in the field. These photos will be analysed by our image recognition algorithm which quantifies the exact plant nutrient content for the macronutrients (N, P, K, S and Mg). The SIITAg system then retrieves a satellite image with relative field nutrient uptake and calibrates the relative values with this in-situ data, generating absolute uptake values for the entire field. Based on the crop type and crop condition, SIITAg will generate a variable rate fertilization map with the variable amounts of nutrients for each zone needed to attain optimum nutrient concentration, which the farmer has the option to tweak if desired. The application map can then be uploaded directly to the farmer’s fertilizer spreader through ISOBUS technology (if available). The functioning of the whole system is outlined in the following figure.
Space Added Value
SIITAg uses earth observation images from Sentinel-2, as well as the Galileo GNSS of ESA for computational purposes. The images from Sentinel-2 enable our program to perform field boundary detection, produce relative nutrient deficiency maps, and analyse field performances. Together, this results in field management zone recommendations for farmers.
Unfortunately, satellite-based nutrition maps only show relative information about deficiencies, not absolute amounts. Therefore, SIITAg calibrates satellite images for each field with in-situ sample pictures of crop leaves. When taking a photo with the app, the phone also connects with the GNSS to provide the geolocation of the picture within the field.
GNSS is also used to generate the variable fertilization map, a machine-readable map for the fertilization machine enabling the distribution of designated amounts of fertilizer at specific geolocated field management zones.
Current methods, such as a mobile app catalogue or field nutrition sensors require canvassing entire fields and taking samples every few metres to achieve the same result. Working with satellite imagery enables SIITAg to cover entire fields at once, which is an important factor in terms of time- and cost-efficiency, and thus in scalability.
The Final Review was successfully held on 10 August 2018 and the study has been completed.
Spacenus has proven the feasibility of SIITAg in March 2018. SIITAg showed impressive test results of the PND service and gathered interested of important players in the agricultural market.
As a next step, we are preparing for the demonstration phase, and have already built up key partnerships to get access to critical market channels. We are planning to begin our demonstration phase from October 2018 to last 24 months