Engineering R&D: Trimming the fat
Controlling the fat content of meat means getting top dollar for each grade and keeping customers pleased that they’re getting what they expect for their money.
FPL Food (Augusta, GA) is one of the largest privately owned processors of fresh beef products and value-added meat selections in the US. It provides several brands of high-quality beef to retail and foodservice customers. The company prides itself on its integrated “pasture to plate” system, whereby it is supplied by its own farms that raise 2,500 head of cattle annually to meet its customers’ needs.
For most meat processors, grading the beef for fat content after it’s been trimmed is not a speedy process; chubs of beef or samples of ground beef are sent to the lab to be checked for lean/fat content. Once the samples have been tested and graded, the processor can mark them as tested and ship them. If they don’t meet the required lean/fat ratio, the processor could elect to retrim or regrind, which means sending them back to the lab to be checked again for an appropriate lean/fat ratio. This can a be hit-or-miss method, and the processor stands to lose out when it comes to pricing meat products optimally and getting them out the door on time.
Tyler Randolph, FPL corporate engineering manager, decided it was time for a change, especially since X-ray analysis equipment is available to make real-time assessments of trimmed and ground meats. With some input from FOSS (which provides online instrumentation) and Friesen’s Inc., a supplier of material handling equipment, Randolph designed a trim management system that produces real-time results of lean/fat ratios and eliminates individual sampling from the equation. FE asked Randolph to provide some more detail on his design.
FE: Why is FPL’s trim management system important to your business?
Tyler Randolph: The system takes a random flow of beef coming from our trim tables and converts that flow into discreet portions of beef at a given weight. Then, it sequences them through an X-ray machine that measures the fat/lean ratio of the portions. Next, the beef portions fall onto a sortaway conveyor that uses a custom fuzzy logic sorting algorithm to determine the sorting locations for the portions to optimize their value and achieve the desired target fat/lean ratios. When a batch meets its target, a label is printed for the combo, and all the data is pushed to our database for traceability and analysis. With this system, we have real-time information about the beef we are boning, and we minimize variation in finished fat/lean ratios by having a very controlled process.
FE: What part is automated, and what isn’t?
Randolph: There is a station where the random flow of beef is manually presorted to produce very lean and very fat portions. These two extremes can then be combined to achieve any fat/lean ratio. The system operates in a 100 percent automatic mode, even if no presorting is done. But for the best results, presorting is required. All the sorting decisions are done automatically based on the specified target ratios.
FE: What differentiates FPL’s system?
Randolph: Typically, in the trim beef industry, a random flow of beef is manually sorted into holding bins of lean and fat beef. Then, one combo is made at a time by pulling the needed fat or lean beef from the bins and running it through an X-ray to hit the target fat/lean ratio. The downside to this is that processors are bound to the average fat/lean ratio of each holding bin and limited in their ability to optimize every piece of the beef’s value. In contrast, the FPL system is not bound by averages and can extract more of the high-value beef that can be sold at a better value. This methodology has been used in trim management before but only for small pieces of meat and low volume. The FPL system handles large pieces of beef [up to 40 lb. each] and high throughput, unlike other systems that require different material handling methodologies.
FE: How do people interface with this system?
Randolph: An operator uses a touchscreen next to the system to designate the target fat/lean ratio and target weight for each of the six loading locations. This screen gives the real-time status of each location, which managers can remotely view from any computer in the plant.
FE: What are the benefits of real-time analysis compared to taking samples to the lab?
Randolph: The biggest benefit is the 100 percent sampling achieved by the inline X-ray machine. This eliminates all the sampling errors traditionally associated with extracting a 5-lb. sample from a 2,000-lb. combo of meat.
FE: How does the accuracy of the inline system compare with that of a lab instrument?
Randolph: Our tests show the X-ray values are much more accurate than the values achieved with lab instruments.
FE: What tools did you use to design FPL’s trim management system?
Randolph: The overall design was modeled in 3D using Solidworks. This model was used in the creation of the system specification documents we sent out to the vendors bidding on the project.
FE: How did suppliers help in the overall design?
Randolph: We partnered with Friesen’s in the development of the finished system. With the FPL 3D model as a starting point, Friesen’s did the detail designs and worked closely with us in all aspects of the design, from drawing approvals in full 3D to developing the sorting algorithm. We worked with FOSS throughout the process to ensure we were using its X-ray equipment in the best possible way.
FE: What might be the takeaway for other processors that are considering a trim management system?
Randolph: The greatest challenge we had to overcome was the inherent randomness of meat. Automation thrives on consistency, and there is nothing consistent with meat. Every piece is different in shape and weight. What makes this system work day after day is its many redundant backups that account for the unknowns that come with meat and the development of new technology. Plus, you need to partner with a team dedicated to excellence.