Artificial intelligence helps coordinate precision agriculture tasks
In the digital era, no sector is left untouched. Even traditional sectors such as farming are becoming increasingly digitised. For example, farmers across Europe are now using technology and data for everything from automatically steering farm machinery to applying fertilisers, tracking livestock, monitoring soil quality and improving crop yields. The problem with this field, which is often referred to as precision agriculture, is that it tends to be rather piecemeal, with separate solutions for every task. But what farmers need are comprehensive yet flexible and easy-to-use solutions capable of automating all these tasks in a coordinated way. Enter the EU-funded FLEXIGROBOTS project. Through the use of such cutting-edge technologies as the internet of things, big data, artificial intelligence (AI) and robotics, the project aims to revolutionise agriculture. “FLEXIGROBOTS addresses the growing need for new robotic technologies and their adoption by ensuring the efficient automation of precision agriculture operations and flexible use of multi-robot systems,” says Raquel Lazcano, R&D team leader at Atos IT Solutions and Services Iberia, the project’s coordinating partner.
One AI platform for multi-robots doing multiple precision agriculture tasks
At the heart of the project is an AI platform for building heterogeneous, multi-robot systems that can collaborate on complex precision agriculture missions. “Our platform aims to enable fleets of heterogeneous robots for precision agricultural tasks, where any technology provider can participate and where farmers will have full control of the data collected from their fields,” explains Lazcano. In doing so, the project aims to bring more versatility to precision farming, allowing the same robots to be used for different tasks. Furthermore, the platform is designed to ensure more cooperation between ground and aerial robots and increase the use of data across the agriculture value chain. The FLEXIGROBOTS platform also looks to bring more autonomy and precision to precision farming-based tasks – all while reducing the level of expertise needed to use such technologies. “By simplifying the technical expertise needed to put AI models into production, we’ve opened the door to creating powerful applications to support farmer needs,” adds Lazcano. Not only does the platform comply with the International Data Spaces Association’s reference on data sharing, it also adheres to the requirements for AI trustworthiness established by the European Commission high-level expert group on AI.
Demonstrating the benefits of AI-driven precision agriculture
The platform has been tested in three real-world, large-scale pilots using different types of crops and in various geographic regions. For example, in Spain, the project demonstrated the high capacity and versatility of robots to carry out different tasks in vineyards, contributing to the economic profitability and quality of grapes used to produce wine. In Finland, researchers used FLEXIGROBOTS’ technologies to manage pests and to guide heavy machinery, while in Lithuania and Serbia they were used to automate such tedious tasks as soil sample extraction and analysis on blueberry farms. “These pilots demonstrate the many benefits of taking a unified, AI-driven approach to precision agriculture,” concludes Lazcano. The project is currently working to further develop its technology, with the aim of providing market-ready applications. For instance, CSIC, one of the project’s partners, is already looking to patent an autonomous robotic assistance system for harvesting. Atos meanwhile is exploring including the AI platform within its product portfolio.
Keywords
FLEXIGROBOTS, artificial intelligence, AI, precision agriculture, agriculture, farming, data
