Beef robotics project moving ahead, but still some way from commercial role + VIDEO

Jon Condon, 22/07/2016

THE development of automated carcase scanning and cutting systems now being adopted in the lamb industry has started in the beef sector, but the process still has a considerable way to go before commercial uptake can happen, feedback suggests.

Beef Central readers may be familiar with earlier articles we’ve published on lamb robotics and scanning applications, like the one included in the video link provided below, at JBS Bordertown, South Australia.

But the sheer size and inconsistency of beef carcases, compared with lamb, is just one of the factors which means the process is a long way from simply substituting one species for another.

JBS’s Dinmore processing plant near Brisbane – the largest beef plant in Australia – is the site of the first development trials for beef in Australia.

The system at Dinmore has been operating on a regular basis for more than 12 months.

The process is using DEXA X-ray technology to scan carcases, plus three-dimensional colour cameras to assess each carcase along the chain, and apply two vertical rib cuts to carcases using robotic saws. The objective of using the assessment data is to make accurate cut placements in order to optimise boning room yield, and delivering greater efficiency.

The system has been trialled on a range of beef carcase types, excluding bulls.

One of the important next steps is to try to extend the assessment capability towards providing accurate meat yields on each carcase. Developers hope that by the end of 2016, the beef system will be ‘nearing’ the performance levels currently reached by the equivalent lamb systems.

Perhaps underpinning the relatively slow start by the beef industry in pushing into this technology is some of the lingering doubts about such systems, that still exist from the days of the brave, but ultimately futile attempts to deliver the Fututech automation project back in the early 1990s.

MLA’s general manager, value-chain innovation, Sean Starling, said he anticipated that beef automation would ‘really start to get exciting’ over the next 12 to 24 months.

“It will be very similar to the lamb approach, in focussing on automating the carcase cutting of those high-value areas of the carcase, for maximum yield,” he told Beef Central.

In the case of lamb, trials have consistently shown a $4-5 per head advantage in yield in using the technology, and the expectation is that will ‘easily be doubled’ in beef carcases.

But the assumption that labour is a big part of the motivation in investing in such technology development is not quite accurate.

“Typically, 80 percent of the benefit in such automation is in improved yield,” Mr Starling said. “Of the remaining 20 percent, maybe one third of that is labour saving. It’s not the main objective,” he said.

“The system only reduces the labour component on those particular manual roles, which are not particularly high in labour content. But they’re often quite dangerous jobs.”

“But we still can’t automate everything. We still need people with those skills, but now they have more time to do a fewer number, and inherently become more accurate in the task they do, in their own right. And it takes people away from more dangerous tasks,” he said.

“Nobody has ever lost their job because of such installations – they just get deployed elsewhere in the plant.”

The technology at Dinmore is designed to deliver up to 440 cuts per hour.

The Dinmore development project, which beef Central understands to have cost well over $1 million in total cost to this point, is an MLA Plant Initiated Project (PIP), receiving funding support from JBS, MLA, Australian Meat Processors Corporation and technology developers, Scott Robotics & Automation.

“It’s still early days, but I think the next five years in the beef space are going to be really exciting,” Mr Starling said.

Still a long way to go in beef development process

JBS Northern COO, Anthony Pratt

JBS Northern COO, Anthony Pratt

JBS Northern chief operating officer Anthony Pratt said it was important to manage expectations about where the technology process was up to.

“This is still very much R&D – we are a long way off, yet, from moving into any real commercial adoption,” he said.

“It’s far-removed from simply plugging-in a bunch of robots and starting to grade carcases on yield,” he said.

“At this point, the process certainly does not function reliably. It takes years and years to perfect it. The system has already been installed for two years, and operating for close to 18 months but it is not without its problems. Breakdowns are common, or the algorithm (software) goes haywire and it starts cutting a carcase to pieces. We still have a considerable way to go.”

While it as going to be a ‘slow burn,’ JBS remain absolutely committed to continuing to work with the system, he said.

“We’re all invested in it, and working hard to get it right, but we think it is going to take years, yet to get it to a point where it is commercially viable.”

“We obviously think the potential is there, otherwise we would not have invested in the project, or the technology company that is driving it (Scott Automation & Robotics),” Mr Pratt said.

“We absolutely believe this will ultimately deliver enormous benefits to us, and the broader industry, but it is going to take time. People have to be realistic in their time expectations,” he said.

Once fully developed and in commercial use, the technology has potential, at least, to plug into MLA’s new ‘Value Chain Digital Strategy’ discussed in this separate story published this morning.





Your email address will not be published. Required fields are marked *

Your comment will not appear until it has been moderated.
Contributions that contravene our Comments Policy will not be published.


  1. David Hill, 22/07/2016

    Beef producers may well be disappointed to hear that the objective of using the assessment data is to make accurate cut placements in order to optimise boning room yield, and delivering greater efficiency. While producers can remember the brave, but ultimately futile attempts to deliver the Fututech automation, they can also remember ViaScan. Obviously for automated cutting the scanning data needs to be accurate, is it the un-reliability of the scanning data or the cutting automation that has this years away? MLA’s general manager,value chain innovation Sean Starling, said he anticipated that beef automation would ‘really start to get exciting over the next 12 to 24 months’.
    As a beef producer I have been waiting to get excited about being objectively paid on carcase yield for around 20 years, I would suggest if Mr Starling believes this all about automation, he should be the innovation general manager for AMPC, which is the processing industries service provider, I thought MLA was supposed to be the service provider for producers foremost?
    It is my understanding that whilst producer levies are only used in ‘proof of concept’ for Plant Initiated Projects, there is tax payer funds used as part of these R&D projects, maybe this should also have been mentioned. Producers are tied of getting paid on P8 fat as an indicator of yield. Objectively measuring yield and quality in carcasses and having returns based on these criteria should be very high on industries priority list, surely the fact that automation technology is not at the same stage as objective scanning shouldn’t stop the implementation of Objective Carcase Measuring as a means of value determination?

Get Beef Central's news headlines emailed to you -