The need for automation in the meat industry is increasing due to high demand, challenging working conditions, and a shortage of skilled labor. New technology is taking automation a significant step forward.
The development of an adaptive robotic system that integrates 3D imaging, CT scans, and artificial intelligence is a major advancement in meat industry automation. This system is designed to handle the natural variability in the anatomy of pig carcasses and represents a new production paradigm in meat factory cells.
The research by NMBU PhD candidate Ian de Medeiros Esper addresses the challenges associated with adaptive cutting and deboning of pig carcasses. The goal is to develop a solution that is both efficient and accurate. To achieve this, de Medeiros Esper has explored various technologies and methods in his doctoral work at NMBU’s Faculty of Science and Technology.
Meat Factory Cells (MFC)
de Medeiros Esper focuses on the automation of pork processing within the Meat Factory Cell (MFC) system. His research presents five contributions that cover various aspects of automation and robotics in meat processing.
- Systematic review of advances
The research provides a systematic review of advances in robotics and automation in slaughterhouses. de Medeiros Esper analyzes commercial products and research publications to assess the current state and examines the adaptability and feasibility of various systems for parallel cell-based production. - 3D reconstruction in meat processing plants
He introduces a method for reconstructing 3D point clouds from cameras placed arbitrarily around a symmetrical object with minimal overlap. This method addresses the challenges in the slaughter line environment and the high degree of symmetry required in pig carcasses. - Open dataset for machine learning in meat processing plants
The third contribution presents an open dataset of pig carcass cutting for applications in computer vision and machine learning. The dataset aims to support the development of automation, object recognition, classification, and semantic segmentation in meat processing plants. - Applications of RGB-D cameras for 3D scene reconstruction
de Medeiros Esper investigates the use of RGB-D cameras for 3D scene reconstruction and evaluates different methods for position estimation. - Framework for automation of modified primal cuts on pigs
Avslutningsvis skisserer han et rammeverk for automatisering av modifiserte primalsnitt på gFinally, he outlines a framework for the automation of modified primal cuts on pigs in an MFC context. The framework integrates CT scans, 3D point clouds, and cutting models to handle anatomical variability and increase precision in meat processing.
– Together, these contributions provide insights into the challenges and opportunities of integrating robotics and computational intelligence in the meat industry. The findings suggest potential areas for future innovation, especially for small and medium-sized producers, says de Medeiros Esper.
Ian de Medeiros Esper will defend his PhD thesis "Intelligent Cutting in Meat Processing: Integrating CT-scans and 3D Imaging with Artificial Intelligence in a Framework to Overcome Natural Variability" on 15 November 2024. Read more about the event here.