Launched with the aim to cut down the use of pesticides and plant protection products in accordance with the EU Green Deal, Smart Droplets will deploy deep tech to produce more quality food with less resources, creating a long- term impact across European agrifood sector.
The conventional spraying methods result in excessive pesticide and fertilizer use, posing severe environmental impact – being responsible for soil, water and air pollution, and harming both humans and ecosystem fauna and flora. Armed with 2,9 million euros from the EU, with a total budget of 3,1 million, the project will create novel environmentally- friendly practices for reduction using autonomous robotic platforms, innovative spraying, digital twin and AI models on two critical cropping systems: wheat and apple. 9 academic and leading industry partners across 6 countries will make sure that within 4 years, all the objectives are achieved.
In the past 9 years, there has been an astronomical rise of toxic pesticide residues on fruits and vegetables in Europe. According to the research by Pesticide Action Network, the frequency of “more hazardous pesticide” residues on fruits is increased by 53%, while vegetables showed a 19% increase from 2011-2019. Moreover, todays spraying operations lack of more selective and targeted approach and do not consider specific characteristics of plants. On many occasions, chemicals are mixed, hindering their efficacy.
This is in line with the EU Green Deal goal that aims for a 50% percent reduction in pesticide use and 20% reduction in fertilizers by 2030.
If this seems easier said than done, that’s because it is. But a new EU pilot project aims to work on it. Smart Droplets’s bid is that the EU Green Deal goals cannot be reached without the use of advanced technologies that will help in pesticide use reduction, low carbon emissions, low water use and food waste reduction.
The project’s slogan is turning less into more in European farming and Smart Droplets’s mission is to transform food production in Europe by testing deeptech enablers on two critical cropping systems of wheat and apple orchards.
“With enabling technologies like robotics, AI digital twins, data infrastructures and direct injection spraying that reduce pesticide and fertilizer use, the project will deliver a holistic system capable of translating large amounts of data into meaningful information and impactful spraying.”
Spyros Fountas, associate professor at the Agricultural University of Athens
In terms of specific use of technologies, Smart Droplets will rely upon established data infrastructures for agriculture purposes and develop the necessary extensions that will host all the foreseen innovation. It will then be tested with existing Farm Management Information Software already in use by farmers.
Smart Droplets will use artificial intelligence models to enable actionable decisions at the field level, about anomaly detection or predicting the evolution of crops and presence of threats. With constant feeding of new training data, the hope is to further enhance accuracy and generalization capacity to new unforeseen events and agriculture environments. The project will bring such models closer to commercialization, by linking the most recent advances in AI in order to increase the prediction accuracy and robustness.
Digital twins will be used to accommodate different sources of agronomic data and utilize process-based models of crops and diseases for simulating field conditions and biological mechanisms. Smart Droplets’ solution will adopt a hybrid approach for spraying recommendations, combining predictions on regional, farm and plant level. It will also be able to flag anomalous situations in which human intervention is required.
Smart Droplets will deploy robotic systems that can operate autonomously to perform spraying tasks with increased spatial awareness. Thanks to an optimal path planning and task-driven objectives, they will be able to navigate effectively. An autonomous retrofit tractor will carry the spraying machine and sensors to execute the spraying tasks successfully.
The direct injection spraying technology will enable the separate application of different chemicals at one task, thus dramatically reducing off-target spraying. This process will further improve this functionality through real-time perception systems and upgraded components on the sprayers, ensuring spraying accuracy for the deployed hardware.
Together it works like this: Field awareness before the tractor and sprayer deployment allow for a probabilistic planning of the spraying and chemical selection based on AI models and the Digital Twin information. This reduces gaps in application, latencies and off-target spraying. The cameras mounted on the sprayer will then validate Digital Twin’s prediction and fine-tune the dosage recommendations in real time. A maximum of 5 chemicals on different states and viscosities will be accommodated in the DIS, claiming savings of 30% in Plant protection Products (PPPs), 15% in nutrients and 30% in overall water use.
For testing this combination of technologies, SmartDroplets will deploy the system in two diversified and complementary environments to evaluate and test its components as thoroughly as possible: Apple orchards in Spain, and the second in a commercial wheat farm in Lithuania.
The crops are chosen for distinct reasons: Apples, as they are among the crops where the highest amounts of pesticides are used, and Wheat accounts for 44% of the main cereal production in the EU-27.
Building a community around ideas means more for everyone. Collaborations with multiple stakeholders and knowledge transfer is an integral part of Smart Droplets’ goals.
The project will ensure proper information exchange with relevant stakeholders, outreach to similar communities and training programs for the community members. Smart Droplets will foster this community building focusing on communication channels, digital and physical platforms for interaction among members.