A team of researchers of the Instituto de Automática e Informática Industrial (Instituto ai2) and the Centro de Investigación en Gestión e Ingeniería de Producción (CIGIP) of the Universitat Politècnica de València (UPV), together with other eight European organisations, are working on the development of IA technologies that, incorporated to autonomous drones, will improve the management of the wild berry picking in Finland and will avoid to lose 90% of production, as today occurs.
Approximately 500 million kg of berries grow naturally in Finland every year. But less than 10% of the wild berry crop is harvested from the country’s forests. The challenge in collecting wild berries lies in manual forest harvesting. Due to the short season, the majority of the work is conducted by foreign workers with limited knowledge of the language, culture, and landscape.
The FEROX project was born with the aim of helping the pickers and make picking process more effective. The 3-years project, funded by the European Union, will employ autonomous drones equipped with various sensors to acquire data, and build 3D models of the forests. These efforts will make it possible to accurately predict the location, amount, and types of berries. The collected data will be used to build AI models to help workers locate berries and optimise their operations. In addition, FEROX will provide wild berry pickers with navigation and locating services and physical support to improve their working conditions. The holistic solution of FEROX will contribute to the overall safety of the workers by automatically monitoring the pickers and providing aid where it is needed.
Drone fleet management
The team of researchers in the Universitat Politècnica de València (UPV), led by Francisco Blanes, will focus both in the technical management of the project and the development of core technologies. Specifically, according to Laura Smith, researcher at Instituto ai2-UPV, “the main tasks we will developed consist on the definition of a system architecture that meets the requirements of the solution proposed within the project; and adapting and implementing a fleet management system previously developed in mobile robots in UPV, and now applied to a drone fleet. In addition, we will also collaborate in tasks such as the development of IA algorithms, and the hardware and software need to communications systems. This point is specially significant within the project, as working with drones in the forests of North Finland introduces several challenges, such as the difficulty of these vehicles to navigation and data transmission under the deep forest, and other technological limitations”.
Nine European partners
Besides UPV, the Ferox team is composed by the following organisations:
Fondazione Bruno Kessler (FBK), Italy – Fondazione Bruno Kessler is an award-winning research institute. FBK specialises in the areas of Information Technology, AI, Sustainable Energy, Sensors and Devices, Cybersecurity, Digital Society, and Digital Industry. Their expertise will inform much of the technical aspects of the project, like machine vision and 3D applications. FBK coordinates the FEROX project.
Deep Forestry, Sweden – Deep Forestry builds drones to automate and digitise commercial forestry fieldwork and tree-based agriculture. Within the FEROX project, their role is to provide drones equipped with cameras and sensors that will collect much-needed data.
Sparkd AI, Ireland – Sparkd helps companies innovate their products with computer vision and deep learning technologies. Sparkd’s role within the FEROX project is to design and develop computer vision and machine learning algorithms. By doing so, Sparkd aims to fulfil three goals; optimising yield, increasing workers’ safety, and increasing the quality of the harvest.
Tampere University (TAU), Finland -Tampere University is a multidisciplinary institution in Finland, focused on research and innovation in technology, health, and society. Within the FEROX project, their areas of focus are the research and dissemination of the solution’s core technologies.
National Land Survey of Finland (FGI), Finland – The National Land Survey of Finland is a 200-year-old organisation that safeguards, monitors, and maps Finnish land. As part of the FEROX project, The National Land Survey of Finland will support research and mapping efforts.
INGENIARIUS, Portugal – INGENIARIUS is an organisation devoted to the development of mobile robotics solutions. Within the FEROX project, INGENIARIUS will spearhead system integration by coordinating the efforts of all the partners. The INGENIARIUS team is also responsible for fleet management and the development of the user interface.
ARKTISET AROMIT – Arctic Flavours Association (AFA), Finland – The Arctic Flavours Association is a Finnish non-wood forest products association focused on the production and maintenance of wild berries, mushrooms, herbs, and special forest products. Their goal is to promote the gathering, processing, and use of natural products while also improving their quality. The AFA will lead the dissemination and demonstration of the FEROX project.
Cranfield University, UK – Cranfield University is a specialist postgraduate institution specialising in the fields of technology and management. Within the FEROX project, CU will conduct human analysis, with the aim of measuring the psychological impact of the project on workers. Cranfield University will also be key to the ethical management of the FEROX project.
Together, these partners can develop and repurpose this technology for industrialised cultivation and supporting global sustainability.
Commenting on the potential of FEROX, project coordinator Paul Chippendale said: “FEROX opens up vast possibilities and there are 3 main areas of impact we have identified: The first is scientific as the project explores the benefits of AI and drones in wilderness applications. Societal; as we think the project will help instill trust in these technologies and finally is economic as we predict new business opportunities for SME’s and citizens”.