Publications

Scientific articles and deliverables from the project

Scientific articles from the project

RoBUTCHER: A novel robotic meat factory cell platform

Alex Mason, Ian Esper, Olga Korostynska, Luis Eduardo Cordova-Lopez, Dmytro Romanov, Michaela Pinceková, Per Håkon Bjørnstad, Ole Alvseike, Anton Popov, Ole Smolkin, Maksym Manko, Lars Bager Christensen, Kritof Takács and Tamas Haidegger

Published in The International Journal of Robotics Research, 21, Feburary 2024

Automation is critically important for sustainability in meat production, where heavy reliance on human labour is a growing challenge. In this work, a novel robotic Meat Factory Cell (MFC) platform presents the opportunity for unconventional automation in pork meat processing, particularly abattoirs. Instead of following line-based approaches, which are the main option today, it uses robotics and Artificial Intelligence (AI) to perform complex cutting and manipulation operations on entire unchilled pork carcasses, with awareness of biological variation and deformation. The long-term goal of the MFC is to take a pork carcass as an input and produce seven primal outputs: hams, shoulders, saddle, belly and entire organ set. However, the MFC platform is under continuous development – therefore, this paper aims to demonstrate it through a specific use-case: shoulder removal. The system is evaluated based on data from testing and development sessions (June–November 2022), with a total of 34 attempted shoulder removals. Data regarding the MFCs’ ability to handle variation, in addition to success rate and process timing models are presented. Qualitative feedback from skilled butchers is also discussed. The authors propose that, as well as technical development of the platform, it is important to consider new ways of comparing unconventional systems with their conventional counterparts. Innovative manufacturing systems have more to offer than raw speed and volume; traits such as flexibility, robustness and scalability – particularly economic scalability – should play a prominent role. Future legislation and standards must also encourage innovation rather than hinder innovative robotics solutions.

3D model based adaptive cutting system for the meat factory cell: Overcoming natural variability

Ian de Medeiros Esper, Lars Erik Gangsei, Luis Eduardo Cordova-Lopez, Dmytro Romanov, Per Håkon Bjørnstad, Ole Alvseike, Pål Johan From & Alex Mason

Published in Smart Agricultural Technology, 7, March 2024

This article presents a comprehensive framework for executing primal cuts on pigs within a Meat Factory Cell (MFC) context, with potential applications for small and medium-sized producers. The framework begins by creating a 3D model from CT-scans, which is then aligned with a 3D point cloud acquired from an Intel© Realsense™ camera using an initial coarse estimate, and refined through Bayesian Coherent Point Drift. Cutting trajectories are generated based on a custom 3D model of the cutting surface, designed with consideration of the pig’s skeletal structure and the cutting properties of the knife tool attached to the robot. A qualitative evaluation of the cuts performed by a professional butcher reveals promising results, while also identifying areas for improvement. The article underscores the potential of integrating CT-scans, 3D point clouds, and cutting models to automate primal cuts in the meat industry, addressing the inherent anatomical variability among animals.

Social performance and impact assessment of an autonomous system in the meat processing sector

Clara Valente, Rannvá Danielsen, Anna Woodhouse, Fredrik Moltu Johnsen & Ellen-Marie Forsberg

Published in The International Journal of Life Cycle Assessment, 7, December 2023

The automation of pork processing through robotics raises important societal concerns regarding working conditions of slaughterhouse workers and impacts on local communities. This article aims to evaluate the social performance and impacts of implementing an Autonomous Robotic System (ARS) for meat processing, comparing pre- and post-implementation scenarios.

Time for Change: The Case of Robotic Food Processing

Alex Mason, Tamas Haidegger and Ole Alvseike

Published in IEEE Robotics & Automation Magazine Volume: 30, Issue: 2, June 2023

Research and development in food processing automation is not a topic one comes across every day in the robotics domain since it is historically a main target for special-purpose machinery. In the agrifood landscape, it is most certainly agritech—that is, technological development at the farm level—that has captured the attention of robotics researchers today.

Quality assessment of fresh meat cuts as a performance indicator of knives specifically adapted for robot-assisted operations

Helle Røer, Olga Korostynska, Frøydis Bjerke, Dmytro Romanov, Luis Eduardo Cordova-Lopez, Alex Mason, Per Håkon Bjørnstad, Torunn Thauland Håseth and Ole Alvseike.

Published in Management Science Letters 13 (2023)

Manual labour in slaughterhouses is hazardous work. Workers suffer from injuries and occupational illnesses resulting from repetitive movements with sharp knives. There is a need for a robotic tool which can perform versatile tasks with a high level of precision. This knife must be able to imitate the same primary cuttings of a professional butcher and produce meat products which are acceptable to the end-market. This paper reports the results of a world-wide assessment of the fresh pork meat cuts as a performance indicator of knives specifically adapted for automated operation.

Pigs: A stepwise RGB-D novel pig carcass cutting dataset

Ian de Medeiros Esper, Luiz Eduardo Cordova-Lopez, Dmytro Romanov, Ole Alvseike, Pål Johan From, Alex Mason

Published in Data in Brief, Vol. 41, April 2022

This paper presents a pig carcass cutting dataset, captured from a bespoke frame structure with 6 Intel® RealSense™ Depth Camera D415 cameras attached, and later recorded from a single camera attached to a robotic arm cycling through the positions previously defined by the frame structure. The data is composed of bags files recorded from the Intel’s SDK, which includes RGB-D data and camera intrinsic parameters for each sensor. In addition, ten JSON files with the transformation matrix for each camera in relation to the left/front camera in the structure are provided, five JSON files for the data recorded with the bespoke frame and five JSON files for the data captured with the robotic arm.

Drivers, opportunities, and challenges of the European risk-based meat safety assurance system

Bojan Blagojevic, Truls Nesbakken, Ole Alvseike, Ivar Vågsholm,  Dragan Antic,  Sophia Johler,  Kurt Houf, Diana Meemken, Ivan Nastasijevic, Madalena Vieira Pinto, Boris Antunovic, Milen Georgiev, Lis Alban

Published in Food Control, Vol. 124, June 2021

The traditional meat safety system has significantly contributed to public health protection throughout the last century. However, it has been recognised that this system suffers many flaws – the main being its limited ability to control the currently most important meat-borne hazards. The European Food Safety Authority evaluated meat inspection in the public health context, prioritised meat-borne hazards and proposed a generic framework for a new, risk-based meat safety assurance system. The proposed system aims to combine a range of preventive and control measures, applied at farms and abattoirs and integrated longitudinally, where official meat inspection is incorporated with producers’ food safety management systems into a coherent whole.

Effects of Meat Factory Cell on pork qualities, sensory characteristics and carcass hygiene: an exploratory study

M. Sødring, T. Thauland Håseth, E. Rasten Brunsdon, P. H. Bjørnstad, R.
Sandnes, O. J. Røtterud, A. Mason, I. de Medeiros Esper, E. Hallenstvedt, P.
Agerup, K. Kåsin, B. Egelandsdal & O. Alvseike

Published in Acta Agriculturae Scandinavica, Section A — Animal Science (2022)

The Meat Factory Cell (MFC) concept restructures the slaughter line into cell stations and merges elements of the slaughter- and primal cutting processes. With the MFC approach, most of the primals are removed prior to evisceration. This study describes the effect of the MFC concept on carcass hygiene, carcass yield, meat quality traits, and sensory characteristics of selected MFC products from trials with the very first pig carcasses processed with the MFC approach. Results show that hygiene of MFC carcasses rivals conventionally slaughtered carcasses. For quality variables and sensory characteristics of selected MFC products, the study shows that the MFC approach will result in products that equal, and in some cases surpass, conventional products, provided that proper processing, packaging and chilling is applied.

Meat safety legislation and its opportunities and hurdles for innovative approaches: A review

Gunvor Elise Nagel-Alne, Emil Murphy, Brittany McCauslin, Sigrun J.Hauge, Dorte Lene Schrøder-Petersen, Janne Holthe, Ole Alvseike.

Published in Food Control 141 (2022)

Innovations are necessary to meet future challenges regarding sustainability, animal welfare, slaughter hygiene, meat safety and quality, not at least for optimal balance between these dimensions. The red meat safety legislation texts from Europe, New Zealand, USA, and global guidelines, were analysed for normative formulations (“how it is or should be done”) that may create non-intentional hurdles to innovation and new technology.

Towards human-robot collaboration in meat processing: Challenges and possibilities

Dmytro Romanov, Olga Korostynska, Odd Ivar Lekang, Alex Mason

Published in Journal of Food Engineering vol. 331 (2022)

This article critically reviews automation challenges for robotic applications in the meat industry, among those are heterogeneity of meat pieces and inconsistency of cutting trajectories that must be overcome to achieve the final quality product. It specifically focuses on human-robot collaboration (HRC) that could be applied in the meat industry to address these challenges. The paper elaborates on possible adaptation of HRC in meat industry, based on its achievements in other industries.

Smart knife: technological advances towards smart cutting tools in meat industry automation

A. Mason, D. Romanov, L. E. Cordova-Lopez, O . Korostynska

Published in Sensor Review 42/1 (2022)

Modern meat processing requires automation and robotisation to remain sustainable and adapt to future challenges, including those brought by global infection events. Automation of all or many processes is seen as the way forward, with robots performing various tasks instead of people. Meat cutting is one of these tasks. Smart novel solutions, including smart knives, are required, with the smart knife being able to analyse and predict the meat it cuts. This paper aims to review technologies with the potential to be used as a so-called “smart knife” The criteria for a smart knife are also defined.

Advanced Sensors for Real-Time Monitoring Applications

O. Korostynska, A. Mason

Printed Edition of the Special Issue Published in Sensors (2021)

Water quality is one of the most critical indicators of environmental pollution and it affects
all of us. Water contamination can be accidental or intentional and the consequences are drastic unless the appropriate measures are adopted on the spot. This review provides a critical assessment of the applicability of various technologies for real-time water quality monitoring, focusing on those that have been reportedly tested in real-life scenarios.

A review of unilateral grippers for meat industry automation

S. Ross, O. Korostynska, L.E. Cordova-Lopez, A. Mason 

Published in Trends in Food Science & Technology Volume 119, January 2022

Highlights

  • Overviews current red meat industry challenges that drive the need for automation.
  • Why red meat slaughterhouse automation has not seen widespread automation adoption.
  • Benchmark theoretical holding forces calculated for gripper suitability assessment.
  • Review of unilateral robot grippers for use within a new pig slaughter process.
  • To conclude simple vacuum systems, can potentially manipulate 40 Kg meat pieces.

Robotisation and intelligent systems in abattoirs

Ian de Medeiros Esper, Pål J. From, Alex Mason

Published in Trends in Food Science & Technology 108 (2021)

Highlights

  • Reviewed intelligent robotic systems that are being developed or are already available for the carcass cutting and deboning.
  • Provides an overview of the current relevant automation in abattoirs.
  • Presents Overview, Process and Critical review of research and commercial projects.
  • Commercial products are Frontmatec AiRA Robots, Mayekawa Hamdas-RX, and SCOTT Automated Boning Room.
  • Research are SRDViand Z-cut robotic system, ham deboning system and ECHORD-DEXDEB, and SINTEF GRIBBOT.
  • Evaluate of production efficiency, worker’s health, hygiene standards, scalability to suit all production volumes.
  • Investigate the adaptability of these systems to different production concepts, e.g., cell based factory.

Life cycle sustainability assessment of a novel slaughter concept

Clara Valente, Hanne Møller, Fredrik Moltu Johnsen, Simon Saxegård, Elin Rasten Brunsdon, Ole Arne Alvseike

Published in Journal of Cleaner Production, Volume 272, 1 November 2022

This article presents a Life Cycle Sustainability Assessment of an innovative slaughter concept, i.e., a semi-automated version of the Meat Factory Cell (MFC). The system is characterised by division of labour with close human-robot interaction, as compared to a Conventional Slaughter and Cutting Process (CSCP). A case study is built which considers the conditions at a Norwegian slaughter facility. Several assumptions are made for the MFC as the concept is still in the development phase. A sensitivity analysis has been employed to highlight the key factors leading to changes in the environmental, economic and social aspects of the Life Cycle Sustainability Assessment framework.

Conference proceedings

Curve fitting-based deformation tracking for vision-based robotic applications

Abhaya Pal Singh, Dmytro Romanov, Ekrem Misimi & Alex Mason

Presented at the 11th International Conference on Control, Mechatronics and Automation (ICCMA 2023).

Application of robotics on production lines often involves handling of flexible objects (such as plastic bags containing substance or items of natural origin), which makes it crucial to consider the shape of an item before and after it has been affected
by robotic manipulation. Most of the time deformable items are challenging for the robot in such operations as grasping, cutting, or packaging. The paper addresses issues in tracking object deformation and proposes a solution for deformation tracking to form preliminary knowledge and scene awareness on the robot
side. A curve-fitting-based method was implemented to define the region of interest using pictures acquired from a RealSense D415 camera. The developed approach identifies the maximum number of aligned points and uses it to determine where the deformation occurred. A demonstration video showcasing the proposed approach is available at https://youtu.be/HZ6BcyDFyuc.

Intelligent Cutting System for an Innovative Meat Factory Cell

I. Esper, L. E. Cordova-Lopez, P. J. From, A. Mason

Proceedings of the Challenges in Automated Food Processing workshop at the European Robotics Forum (ERF2021)

This paper presents work relating to an intelligent cutting system for pig carcasses. It generates the cutting trajectories based on the meat factory cell cuts. A 3D point cloud is generated from RGB-D cameras placed arbitrarily in pairs on either side of the pig. The challenge for complete object reconstruction with little or no overlap and a high degree of symmetry is solved using a novel pipeline, then the 3D object is aligned to an atlas of the pig that encompasses the pig’s skin, bones, organs, and the desired cuts.

Human-Robot Collaboration in the Meat Industry

D. Romanov, O. Korostynska, A. Mason

Proceedings of the Challenges in Automated Food Processing workshop at the European Robotics Forum (ERF2021)

An approach to some production steps in the secondary red meat processing can be revised to improve human working conditions and food safety. Some of the meat processing steps are difficult to automate due to the tasks’ nature, but, taking into account the emergence of new technologies, especially in development of collaborative robots and recent advancements in the artificial intelligence (AI), it seems to be possible to solve these challenges in the near future.

Computer Vision for Robot Butcher

Luibivyi, M. Manko, I. Sahumbaiev, O. Smolkin, I. Krashenyi, A. Popov, I. Esper, A. Mason

Proceedings of the Challenges in Automated Food Processing workshop at the European Robotics Forum (ERF2021)

To support the autonomous Meat Factory Cell in which two robots will perform the gripping and cutting of the pig carcass, this work presents the application of the deep learning to locate gripping points on the carcass limbs and prediction of 3D cutting trajectories using U-Net-based approach with ResNet backbones.

Open Issues in Agri-food Robot Standardization—the Red Meat Sector

K. Takacs, A. Mason, L. E. Cordova-Lopez, T. Haidegger

Published in 2021 IEEE 15th International Symposium on Applied Computational Intelligence and Informatics (SACI)

Ensuring the safety of equipment, operator and the environment during robotic operation is paramount. Robotic systems are appearing in more and more professional service applications, while mechanic and control system components are evolving fast themselves, the legislation and standards regarding these topics are lagging behind. In connection with the RoBUTCHER project – which is a pioneer research effort employing industrial robots for completely automated slaughtering – it was revealed that there are no particular standards regulating directly robotics applied to the agri-food application domain. More specifically, the meat industry and the red meat sector within has only seen hygienic standards regarding machinery, not considering human-robot collaboration or safe autonomous robot operation in the abattoirs. The purpose of this paper is to provide a general overview of the relevant standards (and similar guiding documents) that could be used as pathfinders during the development of inherently safe robotic systems. Exploring the standard and legislation landscape should offer some instrumental help regarding the foreseen certification process of meat processing robots and robot cells in the near future.

Robotic grippers for large and soft object manipulation

Kristóf Takács; Alex Mason; Lars Bager Christensen; Tamás Haidegge

Published in 2020 IEEE 20th International Symposium on Computational Intelligence and Informatics (CINTI)

Grasping has always been considered a key domain of cyber-physical systems, through which action physical interaction can be achieved. This paper presents a systematic review of the state-of-the-art robotic soft object gripping solutions aimed for the food-industry, focusing on red meat handling. A categorized analysis about the currently used grippers is provided, that could be used or adapted to robotic meat-processing. The paper enlists various solutions and gripping principles for low-payload applications too, although the emphasis is on the classic shape-locking and force-locking grippers that are potentially capable of grasping and manipulating heavier specimens. The purpose of the scientific literature survey is mainly to identify exceptional and/or remarkable gripper-designs, or completely new gripping concepts, while the patent research presents complete, commercially available solutions

Smart Knife for Robotic Meat Cutting

A. Mason, D. Romanov, L. E. Cordova-Lopez, O. Korostynska

Published in 2021 IEEE Sensors

Automation is a key enabling technology for efficiency improvement in the meat industry. This paper presents the development of a novel smart knife based on radio-and microwave-frequency sensing, which is suitable for automatic robotised cutting tasks. Partial-least-square regression and neural network prediction models are shown to determine contact of the knife with a work object and depth of cut. Using a water model, the knife can predict contact with 1.81% error, and depth with 2.45 mm (± 0.18 mm) mean error. With pork loin, error in contact detection was 2.92%, and mean depth error was 7.22 mm (± 1.39 mm).

Go to full version of the publication

Deliverables

Deliverable: Project Web Portal

ANIMALIA

This document presents the work done and currents status of the RoBUTCHER web portal (robutcher.eu)

Deliverable: Review of Available Cutting Tools: Recommendations regarding direction to develop such tools

ANIMALIA, DTI-DMRI, NMBU, MRI, OBUDA, ROBOTNORGE

Saws, knives and bespoke tools are evaluated for use in robotic slaughtering and primal cutting in an autonomous slaughter cell. As the project is in the initial phase further testing is necessary to make a final decision on the appropriate tools.

Deliverable: Data Management Plan

NMBU

The purpose of the Data Management Plan (DMP) is to provide an analysis of the main elements of the data management policy that will be used in the RoBUTCHER project and by the project Consortium with regard to the project research data.

Deliverable: Gripping State-of-the-Art Review

OBUDA, DTI-DMRI

Applying pulling force is essential in various stages of the pig processing,e.g., when dissecting the limbs or removing the inner organs. Various gripping solutions exist, which may allow for the grasping and gripping of soft tissue organs ranging from skin to bowels. An overview of existing commercial, patented and prototyped version is provided to be able to compare the layout and capabilities of different gripping solutions.

Deliverable: Robotics safety and legislative review

Safety of equipment, operator and the environment during robotic operation is paramount. Robotics is appearing in more and more industrial applications, while robots and robot systems are evolving fast themselves, yet the legislations and standards regarding these topics lagging behind. Since RoBUTCHER is a pioneer project in employing industrial robots for completely automated slaughtering, there is also no particular standard regulating directly this kind of application. The purpose of this document – as the 1st deliverable of T1.2 – is to provide a general overview of the relevant standards (or other official documents) that could be used as guidelines for development, towards a foreseen certification process in the future.

Deliverable: Prototype external gripper

DTI-DMRI

A prototype gripper has been designed, built, and tested on warm and cold carcass limb
gripping with successful results. The stainless‐steel gripper is pneumatic driven with a payload/weight ratio of more than 4 and a self‐weight of 5.2kg. The gripper is designed for cleanability and robustness to handle the limbs without leaving damages to the skin surface. The design principle is a three‐point gripper principle, to cope with anatomical features (Ulna / Radius and Tibia / Fibula) at the preferred gripping position. The shape of the three gripping features is found by experiments on cold products assuming the performance being valid for warm carcasses as well. The validation is made by demonstration of a successful collaboration with a cutting robot in the meat factory cell.

Deliverable: Prototype Internal Gripper

OBUDAUNI

The main task of OBUDAUNI in the project is the development of a smart
gripping tool that can carry out the manipulation/removal of the internal organs.
In the 18 months of the project, the first and second prototype of the gripper
was developed and tested for the pre‐defined gripping‐tasks including the inner
organ removal and limb support.

Deliverable: Cutting tool evaluation

Animalia

The report describes testing and comparing of four knives, the data collection survey and response analysis. All four knives are acceptable when it comes to cutting (surface) quality. The respondents judged all samples acceptable and could not separate robot cut surfaces from
manual cut surfaces of the meat samples.

Deliverable: Student Workshop 1

This deliverable summarises the student workshop organised by OBUDAUNI. It contains brief descriptions of the lectures and invited talks, presents the overall goal and structure of the workshop and contains the attendance sheets too. One of the tasks in T8.1 (Dissemination and communication) was the organisation of a student workshop for international students about the scientific topics occurring in the RoBUTCHER project. The title of the student workshop was RoboSchool – AI and Robotics in the agri-food sector.

Deliverable: Student Workshop 2

This deliverable describes student-level dissemination activities of the RoBUTCHER project, namely the Meat Factory Cell system located at the Norwegian University of Life Sciences, Ås, Norway. Specifically, several student workshop events were held at the University during November 2022, to allow students from several cohorts and backgrounds to experience first-hand the system developed during the project. This is in line with the activities planned in T8.1, Dissemination, exploitation and communication. Four events were arranged, and at least 41 students in Norway and Ukraine, ranging from high school to PhD level, attended.

Deliverable: Report Cell Human Robot Interface

This report describes the Meat Factory Cell (MFC) human – robot interfaces. This includes description and discussion of design of both functional, safety, and hygienic aspects.

Deliverable: Robotics safety and legislative review update

Safety of equipment, operator and the environment during robotic operation is paramount. Robotics is appearing in more and more industrial applications, while robots and robot systems are evolving fast themselves, yet the legislations and standards regarding these topics are lagging. Since RoBUTCHER is a pioneer project in employing industrial robots for completely automated slaughtering, there is also no particular standard regulating directly this kind of application. The purpose of this document – as the second deliverable of T1.2 Safety legislation review – is to provide update of the relevant standards (or other documents) that could be used as guidelines for development, towards a foreseen certification process in the
future.

Deliverable: Adaptive learning and remote intervention

This document presents a detailed description of the release state of Ocellus (BYTEMOTION’s bespoke software package) as a solution for D3.2, i.e., adaptive learning and remote intervention. For the purposes of this deliverable, Ocellus can manage trajectory data provided from the AI system developed by CIKLUM‐UKR with an intuitive, easy to use human to robot interface. VIVE tracker hardware is used as input, while Ocellus now supports the use of externally created tools represented as CAD models in its virtual MFC environment.

Watch video showing override of AI introduced trajectory using manual intervention through VR

Deliverable: Demonstration of Completed System Integration

The purpose of system integration has been to develop a platform which incorporates methods of cutting trajectory prediction, human-robot interfaces, cutting tools and gripping solutions, with the goal of having a complete Meat Factory Cell (MFC) system able to demonstrate cutting and manipulation tasks on entire pig carcasses.

Watch a demonstration of Completed System Integration

Deliverable: Challenges or hurdles related to enduser and authority’s acceptance white paper

Throughout the RoBUTCHER project, a key aspect of the consortium activities is end-users’ engagement. The meat industry has a uniquely close relation to the official meat inspection operated by the competent authorities, who are present controlling every single carcass in the production line. As a consequence, the legislation defines the hygienic quality targets, but also partly instructs how the results shall be met. The legislators need to build trust in new technologies and approaches based on sound scientific principles so that future legislation will facilitate application of optimal technologies.

Deliverable: Report on user engagement activities

This deliverable is a report on user engagement activities. The purpose of the report is to document users’ needs for the RoBUTCHER system based on their interaction with the demonstrator. This is specified T1.5: Focus group on user needs. This report summarises the findings of three focus groups interviews.

End User Workshop 2

This deliverable summarises a series of end-user workshops where the Meat Factory Cell (MFC) system has been demonstrated in a semi-industrial setting to several stakeholders, including meat processing industries, industry associations, universities, researchers, policymakers and future generation meat processing employees. A total of 10 events took place during September 2022 and January – June 2023, with audiences coming mainly from Germany but also from across Europe and as far afield as Japan.