Identification & Background 

Project Name: AcornSelectAi – Acorn Kernel Selection System Using Artificial Intelligence 
 
Project Code: 21144_AcornSelectAI | Call MPr-2023-7 

Main Objective: AcornSelectAi aims to develop a system that proposes an intelligent and automated solution for the separation, analysis and selection of high-quality acorn kernels, offering efficiency, accuracy and sustainability. The application of machine learning (ML), reinforcement learning and cross-domain adaptation algorithms makes AcornSelectAi a robust, scalable and versatile solution, able to adapt to different industrial contexts and products, including other nuts. Multispectral processing technologies, combined with the use of advanced algorithms, allow you to analyze complex parameters such as color, texture, brightness, fluorescence, absorbance, and contaminant identification. This level of analysis ensures a significant reduction in waste and increases the acceptance rate of high-quality kernels, raising the standards of agri-food processing. The AcornSelectAi system also stands out for the implementation of reinforcement learning techniques, which allow continuous improvement with feedback from operators, and for the use of cross-domain learning methodologies, which ensure adaptability to different production conditions.  

Consortium:  
HRV – EQUIPAMENTOS DE PROCESSO, S.A (Líder) 
LANDRATECH, LDA Instituto Politécnico de Bragança 
MORE – LABORATÓRIO COLABORATIVO MONTANHAS DE INVESTIGAÇÃO – ASSOCIAÇÃO  
INSTITUTO DE TELECOMUNICAÇÕES  
UNIVERIDADE DA BEIRA INTERIOR
 
Approval date:29 /05/2025 
Start date: 01/10/2025 
End date: 30/09/2028 

Total Budget: 1.386 684,16 € 
Funding supported:  1.069743,82 € 

The objectives of the AcornSelectAi project are deeply aligned with the impacts which involve both technological innovation and the valorization of a raw material and endogenous resource of high value and great importance, such as acorns.   

  • Precision and Efficiency: Use of advanced technologies to ensure accurate selection and reduced rejection of quality kernels;  
  • Adaptability: Ability to operate efficiently in different contexts industrial and localizations;  
  • Sustainability: Reduction of waste and lower environmental impact, aligning with circular economy goals;  
  • Technological Innovation: Implementation of deep learning methodologies, with models that evolve over time through human feedback;  
  • Scalability: Modular system that can be expanded for new applications and markets. 
  • Activity 1: Survey of Needs, Requirements and Technical Specifications. 
  • Activity 2: Development of the computer vision system for capturing and processing images of the Acorn Kernel. 
  • Activity 3: Development of a system for transport and automatic selection of Acorn Kernels. 
  • Activity 4: Development of the ML system for Acorn Kernel selection. 
  • Activity 5: Integration, Testing and Validation. 
  • Activity 6: Promotion, dissemination and dissemination of results. 
  • Activity 7: Technical Coordination of the Project and Knowledge Management. 
  • ML and Multispectral Computer Vision System for Acorn Core Selection. 
  • AI Algorithms for Classification and Selection of Acorn Kernels. 
  • Dataset of Multispectral Images of Different Species of Acorn kernels. 
  • Intelligent System for Automated Optical Selection of Acorn Kernels. 
  • Intelligent System for Control and Reduction of Contamination.