AUTOMATIC GRADING SYSTEM FOR DETERMINING LEAN-FAT DISTRIBUTION IN PIG CARCASSES

Project Information

PIGSCAN

Grant agreement ID: 312135

Closed project

Start date 1 March 2013

End date 28 February 2015

Funded under

• Specific Programme "Cooperation": Food, Agriculture and Biotechnology

Total cost € 1 333 386,36

EU contribution € 989 319,00

Coordinated by

TEKNOLOGISK INSTITUT
Denmark

Periodic Report Summary 1 - PIGSCAN (AUTOMATIC GRADING SYSTEM FOR DETERMINING LEAN-FAT DISTRIBUTION IN PIG CARCASSES.)

Project Context and Objectives:
Porcine meat, called pig meat by the industry and pork by the consumer, is the most consumed protein in the world (103.2 Mt). In 2010 pig meat production in the EU27 reached 152.6 million head, of which more than half (54.4 %) came from four countries (Germany, Denmark, Spain and France). The EU27 is the second largest pig meat producer (22.9 Mt /year) in the world and is 107% self sufficient and China is the largest producer (47.16 Mt). The pig meat industry is an important contributor to the European economy with a turnover of 66.3 billion €. The industry is very SME intensive (more than 15,000 SMEs comprising breeders, producers, abattoirs, meat processors and packers) and currently employs some 560,000 people.
Pig carcass classification is an important tool for ensuring fair payment to the producer, as well as serving as a tool for contributing to the transparency of the market. In The European Union (EU), special rules are set up for classifying pig carcasses. In 2009 the EC introduced a new regulation (No 1249/2008) which obliges slaughterhouses to classify and weight the animal carcasses. This regulation enables efficient and effective classification and ensures livestock breeders and dealers achieve optimal prices for their meat carcass.
The European classification scheme for grading whole pig carcasses is the S/EUROP system. The value of a pig carcase is determined in particular by its lean-meat content in relation to its weight. Yet still today, in small and medium slaughterhouses the classification is based on human operation, which by its nature represents a labour intensive process.

One the most relevant aspects in pig meat processing is the classification of the 4 primal cuts of the carcass (i.e. ham, loin, belly and shoulder) by their lean-fat content. This information is of utmost importance for selecting the optimum processing for the meat (e.g. drying, curing, cooking, fresh meat, etc). For instance, in the processing of cured hams is crucial to know the lean-fat content (%lean) to adjust the salting time to have a consistent and quality product.
The most widespread system for grading whole pig carcasses is the manual probe (e.g. CGM, Fat-O-Meater, etc) but do not provide information of the primal cuts. Recent studies have demonstrated that the total market value of the carcass is more affected by the individual contribution of each cut (loin, ham, belly and shoulder) rather than the total %lean of the carcass. These individual contributions are dependent on their monetary value on the market, their weight and their lean-fat composition. Existing automatic carcass grading systems provides information of the primal cuts but are still not economically affordable for a significant number of small and medium slaughterhouses.
The PIGSCAN technology will be able to provide an individual characterisation of specific regions of the carcass (i.e. primal cuts). The automatic sorting of the 4 primal cuts still in the whole carcass will allow small/medium slaughterhouses to optimize their production and to ensure that the raw meat is processed efficiently. Manufacturers of the PIGSCAN machine expects to provide the system at a very competitive price compared to current automatic grading systems.

Project Results:

WP1 Magnetic Induction Module

Task 1.1: Different coil geometries and coil configurations tested. The design simplifies manufacturing process allowing fast replication, good repeatability and improved robustness against vibrations.
Dedicated instrumentation designed to achieve a speed of 900 c/hour. The instrumentation has been manufactured in a single electronic board thus simplifying the number of connectors and cabling.
Study on different potential shield materials including Aluminium, Copper and Mu-metal to replace expensive model of ferrite tile.

Task 1.2: Communication interface module based on a PC and an Ethernet connection. PC communication with MI and VIA systems, RFID antenna to obtain carcass ID and weight from the scale. Control unit based on PC for high level operations and synchronisation. Self-diagnostic module (MI module) and remote access through a TCP/IP connection.

Task 1.3 Mechanical structure (developed by GM) configurable to scan carcass in horizontal position and overhead conveyor to allow the automatic pass of the carcass at IRTA facilities. A second simplified mechanical structure (“prototype I” by LENZ) to perform quick tests at several slaughterhouses. Test rig for the VIA system including background panel, lighting system and camera fixation setup (LENZ).

Main results achieved in WP1 (by month 12)
Development of two configurable Magnetic Induction modules. Implementation of different coil geometries and coil configurations. Double emitter configuration provides a more uniform and confined magnetic field distribution. Coil sensors designed for efficient and fast replication. Instrumentation capable to scan up to 900 c/hour. Determination of the best EM shield, based on a cheap ferrite tile model plus an Aluminium back panel. Mechanical design and implementation of the Digital Image Analysis test rig for its use in WP2.

WP2 Digital Image Analysis Module
Task 2.1: Implementation of a VIA system based on a single colour camera and a 3D sensor. Collection of a large set of images at slaughterhouse to make test with the colour camera.

Task 2.2: Development of an algorithm to automatically extract information on carcass dimensions including total length, maximum width and fat depth at the mid line of the carcass. Calibration of the system at lab scale using a reference pattern. The algorithm has been tested off-line using a set of images acquired on-line (700 half c/hour) at slaughterhouse.

Task 2.3: Prototype assembly and test (Task leader: G&M)
A complete laboratory test-rig has been designed and assembled at IRTA by LENZ. The test rig includes all the components (camera, lighting system, computer, etc.) required to determine different dimensions of the pig carcass.

Main achievements
Implementation of a VIA system and calibration at laboratory scale. Collection of images at slaughterhouse (700 halves/hour). Fast algorithm that determines of carcass dimensions including carcass length, maximum width, ham thickness and fat depth (ZP method) in only 15 milliseconds. The proposed 3D sensor avoids the need to turn carcass 90º to determine ham thickness.

WP7 Innovation Related Activities
Task 7.1: Preparation of a project logo, leaflet, publication in magazine, direct communication with potential customers, public website, PIGSCAN project office, reference on the website of partners.

Task 7.2: KM seminar coached by LENZ at M3 and by DTI at M6. Tech watch activities: Description of current SoA, new grading technologies and relevant research carried out by other groups. Market watch activities: Definition of potential markets and potential business opportunities and threats. Preparation of a patent for protecting results achieved in WP1.

Task 7.3: Preparation of a draft plan for the use and dissemination of the foreground (dlv 7.2).

WP8 Consortium management
Activities related with Management of resources including technical, financial and administrative management.

Potential Impact:
Expected final results
The application of an innovative Magnetic Induction system as an alternative to conventional methods for evaluating meat conductivity, avoiding measurement artefacts related to the contact resistance of the electrodes or limitations due to the need to introduce the object inside a chamber or a tunnel.
The development of a simple and low cost Digital Image Analysis module to determine 2D information of the pig carcass.
An automatic system that will be able to determine the total lean content in the carcass (S/EUROP grading).
An automatic system that provides information of the lean-fat content in the four primal cuts.
The development of a cost affordable system. It is expected to achieve a manufacturing target cost for a complete PIGSCAN system below €40,000.

Use
It is expected that the market niche of PIGSCAN will be small/medium pig slaughterhouses using hand held probes. PIGSCAN will solve an important barrier of current technologies. Hand held probes requires an operator and do not provide vital information of prime cuts. On the other hand, 3D ultrasound scanners or VIA system are still expensive for a significant number of pig slaughterhouses. PIGSCAN will provide lean-fat information of prime cuts enabling slaughterhouses to optimize their production and segmenting their production according to specific market requirements.
PIGSCAN will be easy to adopt by the majority of end users given that the price will be affordable even for small pig processing facilities, do not require operator and do not require significant modification of the processing line. The expected commercial price of PIGSCAN will be very competitive when is compared with a cheap hand held probe that cost about €15,000 plus the salary of an operator or when is compared with automatic systems working at similar speeds based in ultrasound (€450,000) or artificial vision (>€180,000).

Socio-economic Impact
The PIGSCAN technology here proposed will result in significant economic benefits to EU pig meat processing companies, who will be benefit from a technology that will let them to accurately classify their pig carcasses according to S/EUROP scheme and extract essential information of the four primal cuts of the carcass for making a more efficient use of them.
PIGSCAN will contribute to satisfy the growing interest in the meat sector to link the carcass value to the real market value of the four primal cuts. A pig farmer typically gets a higher price for a carcass with high %lean. However, high %lean carcasses may provide hams that are optimum for a cooking process but probably not for a cured product.
Through different interviews made with ham producers during the execution of this project, having information about the % lean in the ham (before processing) may lead to a potential increase of benefits in the range of 5-7% per ton of processed product.

Wider societal implications
The new instrument will contribute to job satisfaction and conditions by reducing the stress associated with human dependency and the inherent possibility of human error, as in some instances staff (or independent meat grading experts) may have to evaluate large numbers of live animals or carcasses over a short period of time.
The contribution to improving employment will arise from creating jobs for younger, inexperienced people. Due to the conventional evaluation systems currently being used, only staff with lengthy and professional work experience can carry out carcass classification tasks. However, the easy-to-use apparatus developed during this research project will serve to improve the employment situation especially in the sector of lower skilled labour and in rural areas. It is important to stress that the implementation of this project will not lead to any loss of jobs by replacing manpower with machines, rather it will contribute to saving existing ones.

List of Websites:

www.pigscan.eu