Protect Agri

ResearchProjects Hits: 24 3 minutes read

PROTECT-Agri: Precision Robotics for Efficient Agriculture 

A grant of  Romanian Ministry of Education and Research in National Research, Development and Innovation Plan for 2022-2027, PNCDI IV project no PN-IV-P7-7.1-PED-2024-0571 in Innovation Partnerships Program/ Competitiveness Partnerships Sub-program (PED 2024)” (UEFISCDI), project period: June 2025- June 2027

PROTECT-Agri project aims to develop, demonstrate, and evaluate a multi-purpose autonomous ground robotic platform capable of performing measurements on an individual plant basis. With highly accurate insights at the individual plant level, farmers will be able to respond to the fluctuating availability of growth factors in a more timely and informed manner. The ground plan of a greenhouse does not change over time, so an end-to-end solution specifically for greenhouses will be created and tested. The ground robotic platform will incorporate computer vision methods for vision plant recognition, soil sensing and multispectral plant analysis for plant health analysis. Each crop is spatially registered at the first robotic scouting step by being assigned a crop-ID. Individual crop-IDs are identified during time-consecutive scouting sessions through visual matching, allowing historical records to be generated on a plant-by-plant basis. The beneficiaries will be able to view the outcomes of the robotic measurements via a digital platform that combines information from multiple robotics components and from already-existing Internet of Things (IoT) devices embedded into the greenhouses.

Project Objectives

To ensure the feasibility and the success of the project in and in the proximity of smart cities, the project implies 4 main scientific objectives (SO):

  • SO1 consists in developing a computer-vision ultra-precise plant classification capabilities, integrated into the ground robotic module. Corresponding KPI include ultra-precise data characterizing individual plants in a greenhouse throughout an extended growth phase, achieving a minimum 10% precision enhancement compared to the literature baseline.

  • SO2 refers to the development and incorporation of a ground robotic module with a mobile soil analysis toolkit for in-situ on-board measurements of soil suitability, nutrients, moisture, and minerals, considering individual plant locations. KPI is a location-specific soil characterization data, conveying a rich set of information on soil conditions.

  • SO3 implies the develop a ground robotic module with a plant-level multispectral image analysis feature, allowing precise spatial-temporal characterization of individual plants and fruits. Testing and previous developments in SO1 and SO2 are used in this step. Performance indicators refer to characterization at plant and fruit level to provide specifics in terms of leaf-area-ratio, leaf-fruit-ratio, covering area, and canopy density.

  • SO4 considers the built-up of a robotic platform with seamless integration of computer vision and multispectral sensing capabilities, capable of transmitting data to the central command and control center for aggregation by the digital platform. The autonomous robotic platform is designed for deployment in greenhouses.

  • SO5 refers to development of the digital platform with robot interaction, data management, and analysis built on open standards (establishing middleware for seamless interoperability and integration of all data sources). Resources developed in SO1- SO4 will be optimized and outdoor tested.

    Consortium and Partners


    The National University of Science and Technology Politehnica Bucharest (UNSTPB)
     
    BEAM INNOVATION SRL