Improving Europe’s precision agriculture with AI
Artificial intelligence (AI) has a vast untapped potential in precision agriculture. Aided by remote sensing, Earth Observation (EO) systems, aerial surveys and innovative sensors, precision agriculture can better redistribute and save valuable resources and increase productivity – bringing both economic and ecological benefits. “AI also supports predictive analytics which can, for example, adjust watering with respect to forecasted rainfalls. It can also assess the risks of pest and weed occurrence with respect to dynamic climatic conditions,” explains Giuseppe Salvatore Vella, project coordinator from the Border Management and Defence research unit at Engineering Ingegneria Informatica in Italy. In the EU-funded AgriBIT project, researchers developed and tested a series of new precision agriculture services harnessing AI, to improve Europe’s agriculture value chain. “The project successfully validated the possibilities of combining such AI technology with the precise positioning of remote sensing (satellites and on-field sensors), for near-real-time detection of pest infestations and bacterial infections on crops like industrial tomatoes,” says Piero Scrima, technical leader of the AgriBIT Platform at Engineering Ingegneria Informatica.
Precision agriculture services on a mobile app
Overall, the AgriBIT team released eight precision agriculture services, combining expert agricultural knowledge with EO systems, and on-the-ground technologies including advanced sensors and AI. The services were offered via a centralised web platform and a mobile application, which allows users to collect information from remote sensors. They can use this to monitor the past and current status of their crop fields and the weather status, and predict the future evolution of both. An essential component that added significant value to the prototypes was the inclusion of a cheap GNSS receiver to give satellite positioning. “This provided a more accurate assessment of crop growth, allowing heavy machinery to be used in the field with unprecedented accuracy using georeferenced maps, thus saving valuable space for crop cultivation,” adds Vella.
Piloting phase to evaluate the technologies
AgriBIT ran a 20-month piloting phase, to demonstrate the applicability of these technologies and create a network of services and professionals supporting agriculture across Europe. “The pilots showed that implementing these technologies could be done on any scale and can support a wide range of crops, including vineyard monitoring in Italy, tomato production in Portugal, and peach trees in Greece,” remarks Vella. The technologies generated a noticeable improvement in terms of area monitoring and potential improved quality of harvests by reducing fertilisation, optimising water resources and reducing pesticides. “Technologies have been evaluated on-site on multiple tests, with a variety of sensors deployed and continuously operating even after the project has finished,” adds Vella.
Boosting European precision agriculture
The AgriBIT team hopes the project will popularise precision agriculture within the European Union, eventually improving agricultural yields while saving energy and reducing agriculture’s ecological impact. “Novelties developed in the project are likely to create long-lasting market impact, while having proven also the enormous added value from combining terrestrial sensing with aerial and EO and other satellite-based services,” says Vella. The researchers are now working towards including additional services and improving the AgriBIT precision agriculture platform to become even more flexible and inclusive. They continue to expand the community of users, beyond those who participated in the project. New proposals are being prepared to further advance the technologies, including developing the AgriBIT Machine Learning service, which they hope to bring to European and global markets soon.
Keywords
AgriBIT, precision agriculture, AI, Earth Observation, farming, evaluate, technologies, crops