Amid concerns about drought and energy costs, a team from Andalusia has put forward a tool that allows users to know How much water does a crop really need each day? and adjust the irrigation accordingly, without relying so much on intuition or fixed schedules.
Is AquaCrop-IoT, smart irrigation platform which combines daily images of the plot, weather data and an agronomic simulation model to make recommendations to the farmer optimal irrigation from your mobile deviceTested on a wheat farm in Córdoba, it has managed to cut around a 32% of the water applied without reducing crop yield.
A daily irrigation platform designed for the Andalusian countryside
The development of AquaCrop-IoT stems from a project funded by the Ministry of Universities, Research and Innovation of the Regional Government of Andalusia, with an eye on to make more efficient use of irrigation water in a context of increasing water stress in Spain.
The tool has been designed by a group of Department of Agronomy, University of Córdoba and the Institute for Sustainable Agriculture (IAS-CSIC)Their approach is simple to explain, although technologically complex: to offer daily watering recommendations based on real crop and climate data, accessible from a web application Accessible on any device with a browser.
In this way, a producer can log in from their phone, tablet, or computer and see what volume of water to apply that dayThe platform aims to track the evolution of the plantation and the impact of advancing or delaying irrigation. It seeks to adapt to the real-world pace of a farm, where decisions are made quickly and often with incomplete information.
The underlying objective is twofold: save water and energy without sacrificing production, and offer a tool that is sufficient intuitive so that any farmer can integrate it into their routine without needing extensive technical knowledge.
Digital twin: the virtual replica that decides the irrigation
The core of AquaCrop-IoT is based on the concept of digital twina virtual replica of the plot that attempts to reflect it in detail the state of the crop and its water consumption at any given time. Based on this digital copy, the system simulates different irrigation scenarios and proposes the most efficient option.
As a scientific basis, the platform starts from AquaCrop, the simulation model developed by the Food and Agriculture Organization of the United Nations (FAO)This model, widely used for planning irrigation strategies in herbaceous crops, estimates how the crop responds to different water availability, but was originally designed for a more static use. not to be updated daily with what happens on the farm.
AquaCrop-IoT automates and expands that FAO model, integrating it with physical sensors e near real-time imagesThe idea is that the digital twin should not remain a theoretical projection, but should be continuously corrected by what happens on the ground.
As the IAS-CSIC team points out, AquaCrop simplifies reality and does not address in detail processes such as the appearance of pests, diseases, or nutritional problems. Thanks to the connection with cameras and sensors, the platform allows that The natural evolution of the crop will correct the model when the plant does not behave as expected.
Daily images, sensors, and automatic correction of recommendations
In field operations, a conventional camera installed on the property It takes a picture of the crop every day. From that photograph, the system automatically calculates the plant coverThat is, what percentage of the soil is covered by the leaves. This indicator is closely linked to crop growth and transpiration.
If the platform detects, from the images, that development is slowing down or deviating from what is expected, it can attribute it to situations of water stress, presence of pests, or nutritional deficiencyWith that information, the digital twin is readjusted and the Irrigation recommendations are modified to avoid adding water that would not result in increased yield.
In addition to the images, the system integrates a weather station with twelve sensors distributed across the farm. Among other parameters, it records the solar radiationair temperature, relative humidity, precipitation and wind speed and direction, key variables for estimating evapotranspiration and water balance.
Along with these local measurements, the platform incorporates historical series and forecasts from the State Meteorological Agency (Aemet)Forecasts allow us to anticipate rain events, heat waves, or sudden temperature changes, influencing decisions about when it is advisable to... to bring forward, reduce or delay an irrigation.
According to the team, this entire dataset feeds into the AquaCrop model and updates the digital twin, which becomes a kind of control panel where the farmer can view images, graphs, and projections of crop evolution under different irrigation scenarios.
Meteorology, edge computing, and web access from any device
One of the critical points for a platform of this type to work outside the laboratory is the connectivity on the farmMany farms in Spain, and especially in rural areas, have uneven coveragewhich complicates continuously sending large volumes of data to the cloud.
To overcome this problem, AquaCrop-IoT resorts to the edge computingInstead of relying on a permanent internet connection, the system processes the data directly in the field itself, En a small server installed on the farmand only sends the information necessary for visualization and analysis to the web platform.
This architecture reduces vulnerability to network outages and improves system robustnessThis is essential when it comes to daily irrigation decisions. The farmer can continue using the tool even if the connection is intermittent or experiences occasional outages.
The processed information is then entered into a web interface which aims to be clear and easy to use. It can be accessed from any device with a browser. Updated crop images, key variable charts, and personalized irrigation recommendations, which adjust automatically as conditions change.
As summarized by the researcher from the University of Cordoba Francis PuigFarmers have been consulting the for years climate predictions to manage its exploitation; what this platform now adds is a dynamic simulation of crop evolution and the recommended irrigation, with the aim of cut down on water and energy consumption without harming production.
Durum wheat trial in Cordoba: three irrigations instead of four
To validate the platform's functionality, the team conducted a test in a durum wheat cultivation sown in January 2023 on a private farm in the province of CordobaThe trial coincided with a particularly dry year, a context representative of the water pressure faced by much of agriculture in Andalusia.
During the crop cycle, the digital twin corrected the initial predictions of the AquaCrop model by verifying, for example, that the Plant emergence occurred later than had been estimated. That seemingly minor time difference had an impact on the recommended irrigation schedule.
If only the static model had been followed, AquaCrop would have suggested four waterings with a total volume of 64,8 mm of applied water. However, considering the information from the cameras, sensors, and weather forecasts, AquaCrop-IoT recommended three waterings with a total of 44,1 mm.
The comparison between both strategies results in a saving approximately 32% of the water used for irrigation, without registering appreciable differences in the final crop yieldIn other words, less water was applied and the harvest remained at the same level.
In addition to water usage, the reduction in the number of irrigations implied a reduction in energy consumption associated with the commissioning of the pumping systems. The effect, therefore, was noticeable both environmentally and in the economic cost of the campaign.
Application to other crops, costs and future developments
Although the pilot study focused on durum wheat, researchers insist that the AquaCrop-IoT's approach is transferable to other herbaceous crops, such as corn or vegetables, which respond similarly to irrigation strategies based on the relationship between canopy growth and water availability.
One of the aspects that the team emphasizes the most is the relatively low implementation costThe devices used in the test—mainly cameras and the local processing server—cost around 150 euros per camera and less than 200 euros for the serverFigures that, according to the authors, make them viable for medium-sized farms.
Technical architecture is designed to be modular and scalableThe system is ready to incorporate new sources of information, such as additional soil moisture sensors or images captured by droneswhich would allow for a more detailed and flexible view of the state of the plot.
Looking ahead, those in charge of the project are already working on integrating the platform with mobile apps that allow the farmer take a simple photo with your phone and that the system can automatically calibrate crop growth from that image, without the need to install fixed cameras.
In the words of the UCO researcher Juan Antonio RodrÃguez-DÃazThe agricultural sector is currently immersed in a technological revolution where the sensors already generate a large amount of data, but often They are not integrated into truly useful toolsAquaCrop-IoT's strategy is precisely to transform this avalanche of information into a practical platform that helps to make better-informed irrigation decisions.
All this work positions AquaCrop-IoT as an example of Digitalization applied to irrigation in SpainAt a time when saving water and energy is a priority for farms, the platform demonstrates that it is possible by integrating digital twins, daily images, weather data, and edge computing into a single mobile-accessible tool. reduce the recommended irrigation water by about a third maintaining production, something especially relevant for herbaceous crops in European regions exposed to drought.
