To date in 2019, the Department of Agriculture, Food and Marine has conducted 499 unannounced inspections in 32 beef factories in relation to processors’ carcass grading.

Minister for Agriculture, Food and the Marine Michael Creed confirmed the figure to independent TD for Galway-Roscommon Michael Fitzmaurice following a parliamentary question from the deputy on the matter of inspections and breaches in regulations.

Continuing, the minister noted that, during these inspections, classification officers monitor carcass classification, weights and trim.

To date, one factory has been fined for excess carcass trimming of three carcasses.

“It is intended to publish the names of all factories that have been fined in 2019 on the department website in due course,” Minister Creed added.

“While the classification of beef is compulsory under EU legislation, sheep classification is not mandatory by EU legislation and is implemented as an industry initiative,” the minister concluded.

Grading trial

Meanwhile, the future application of new grading machines modified with more up-to-date technologies “will result in a better performance of carcass classification”, according to the Department of Agriculture, Food and the Marine trial on grading.

A report of the trial – which was conducted during 2018 and 2019 – was published recently by the department, with results long-awaited on the matter.

The trial was organised at Slaney Foods in Bunclody, Co. Wexford. In the summer of 2018, classification results from 2,431 carcasses were used to calibrate the new technology, while in February 2019, a further 2,100 carcasses were used to validate this.

In Bunclody, two machines were used: the existing classification machine; and a second machine with modified technology – a new digital camera and LED lights.

In the trial results, for the 2,100 cattle sample, the modified machine had a total accuracy score of 93.3% for fat – some 7.6% higher than the 85.7% score achieved by the old machine, based on 2002 equations.

Meanwhile, for conformation on the same sample and equations, the modified machine received a score of 85.2%; some 1.1% higher than the 84.1% score given to the current machine.

According to Lindeloof, the accuracy and bias results for the modified machine are at levels greater than those for the classification machines currently in use although the current machine also performed at very high levels.