Toward Better Beef
A fast, accurate, objective method of grading beef?
Once a futuristic scenario, a new wave of technology is rapidly bringing that prospect to reality
by Dan Murphy, contributing editor
For Chuck Akin, it's another day and another shift at Excel Corp.'s beef plant in Dodge City, Kan., located on the state's western prairies in the heart of cattle country. But the nine-year meatpacking veteran is performing a new job today, one that may someday revolutionize beef packing.
From 7 a.m. to 3:30 p.m. each workday, Akin manipulates a bulky camera frame to obtain a video image of the rib eye on individual beef carcasses as they roll by at some 400 cattle an hour.
The video imaging system to which his camera is attached then generates grading information electronically in seconds-a task that has challenged the best-trained human graders since USDA first instituted its system of yield grades almost 80 years ago.
Excel's system is known as Video Image Analysis (VIA), and it's one of several cutting-edge technologies putting a high-tech stamp on the process of grading and sorting-and ultimately improving-the quality of U.S. beef.
"We need a better system to tell us how much each carcass is really worth," says Gary Smith, professor of meat science at Colorado State University, and one of the principal researchers in a comprehensive study to investigate new grading technologies. "When we can predict better, more accurate yield data, then we can send the right signals to producers to grow more marketable cattle."
Toward that end, the National Cattlemen's Association initiated a study in 1994 to assess beef grading instruments. The project was then turned over to the National Live Stock and Meat Board, which provided funding for a study using researchers at Oklahoma State and Colorado State universities.
Working at Excel's Dodge City beef plant in collaboration with Dell Allen, Excel's vice president of quality and training, the researchers are spending five weeks this summer carefully fabricating 100 carcasses to correlate actual yield data with information generated by three types of VIA systems.
The goal is to refine VIA technology to sort carcasses on the basis of boneless boxed beef yield in a closely trimmed program. The study hopes to show that technology such as VIA can move grading forward beyond USDA's five arbitrary Yield Grades, according to Glen Dolezal, professor of animal science at Oklahoma State University and one of the key researchers.
"We're trying to generate yield data independent of USDA," Dolezal says. "We want to be able to tell the packer exactly what percent yield it can expect from each carcass at cutout. If we can consistently do that, then we can slice the pie a lot thinner, and with a lot less variability."
Also being tested in the study are two other technologies: one offering a rapid measurement of total lean content, and the other a way to predict tenderness and eating quality.
Here is a look at the key technologies for beef instrument grading now under investigation.
Video Image Analysis
VIA has been around for years, but the technology first made a splash in manufacturing during the 1980s with the advent of robotic assemblers and quality assurance units in the automotive industry. Powered by a machine vision system, these devices were engineered to "look" at automobile components and make various quality determinations.
"In meatpacking, these video systems go back many years," Allen notes. "As the technology has continued to improve, we're now able to use VIA to help us better utilize each carcass. We're trying to get sophisticated enough that each carcass can be fabricated in optimum fashion."
VIA incorporates a video system-either single or twin cameras- that collect images from both the entire carcass and the exposed rib eye between the 12th and 13th ribs. The images are combined and processed by computer to obtain carcass assessments of fat thickness and rib eye area and assign a Yield Grade.
The VIA systems under study are calibrated to produce yield data. But video systems also hold promise for eventually providing quality data, such as marbling. Video cameras can "see" the marbling pattern in a carcass, for example, and researchers hope to develop an equation that could correlate those images with a quality grade.
"We've now got cameras with a resolution of 1,056 by 1,056 pixels, vs. just 256 pixels only a few years ago," Dolezal notes. "With the higher quality images on the newer color cameras, we're a lot closer to being able to assign marbling scores that will accurately match USDA's quality grades."
Dolezal says work on video assessment of marbling is "still developmental," but points out that a recent study showed an 80 percent correlation between video-generated marbling scores on various cuts of steak and the actual USDA grade.
Whatever its eventual function, one key to optimum utilization of a VIA system is the packer's ability to track individual carcass data through fabrication, Allen stresses.
"You have to be able to collect information on every carcass," he says. "Then, you've got some data you can feed back to the producers who want to work with packers to improve the marketability of their animals."
Originally, Allen says the impetus for development of a video grading system was to provide a tool to assist graders in making their decisions more accurate. At the speed at which today's beef plants run, he recognizes that government graders must make subjective decisions in assigning yield grades.
"They can only estimate the size of the rib eye," he points out. "VIA can quickly and reliably perform that calculation, so it could eventually aid the grader, who could have a monitor at the grading station to display actual rib eye surface area and fat thickness." (See story, page 94.)
By mid-1996, Allen says Excel plans to have a VIA system installed and operating in each of its seven beef plants. But he says for now the video systems will be set up to help make internal decisions about carcass cutout, rather than replacing or augmenting USDA graders.
ToBEC
Based on technology first tried in the 1970s, Total Body Electrical Conductivity (ToBEC) uses an electromagnetically charged tunnel through which primal cuts are conveyored. The unit measures the meat's electrical conductivity and calculates an estimate of total lean meat content.
"Muscle is simply a better conductor of electrical energy than fat," explains Chris Calkins, professor of animal science at the University of Nebraska, and one of the main researchers exploring ToBEC's role in beef grading. "As the beef primal passes through the tunnel, you get a differential reading, or 'conductivity curve' that a computer uses to calculate the pounds of lean in the primal. That number is then used to extrapolate total carcass lean."
Two U.S. pork plants, one in Iowa and another in North Carolina, are currently using ToBEC on-line to evaluate pork carcasses.
ToBEC technology is also being used commercially at more than a dozen beef plants in Australia for determining percent lean in beef trimmings. Boxed product runs through the ToBEC tunnel in about three seconds, and a terminal downstream prints a bar-coded box label identifying percent lean within 1.5 percent accuracy.
Calkins' team is running beef hindquarters through a similar tunnel at Excel's Dodge City plant. "The unit can take 80 readings a second," he says. "The limitation is how fast you can move a piece of beef through the tunnel."
But speed isn't the only limitation.
"The real challenge is whether ToBEC can be adapted for use on the rail," argues James O. Reagan, director of product technology research at the National Live Stock and Meat Board. "We know the technology works on primals. Now we need to develop a system through which the entire carcass can pass. That's crucial to commercial application."
How does ToBEC fit with other technologies, such as video imaging? Calkins says they're complementary, but suggests that if it can be fully refined, ToBEC technology alone is capable of delivering the yield information packers need to know. ToBEC also has application with hot fat trimming, where video imaging would be limited, Calkins points out.
"ToBEC is more accurate than the current grading system, where a grader only has about 13 seconds to determine both yield and quality grades," Calkins notes. "Our studies show that ToBEC data has been at least as accurate-and sometimes more so-in every single instance where carcasses were hand-dissected to verify final yield. It's a technology with outstanding accuracy-and no human error."
Elastography
A promising medical research tool for evaluation of breast cancer, elastography uses ultrasound to measure changes in tissue elasticity following very slight compression. The resulting elastograms offer a cross-sectional image of the tissue sample that gives physicians helpful information on location, size and configuration of suspected tumors.
In medical practice, elastography can help distinguish between hardened tumors and softer, normal breast tissue.
"It's a differential technique," explains Jonathan Ophir, the researcher at the University of Texas Medical School in Houston who pioneered the technology.
"Soft tissue areas compress more; harder ones less. You get a two-dimensional image of the variations in a tissue sample that help physicians locate and evaluate tumors," he adds.
In beef packing, the technology may serve as a non-invasive method to evaluate meat quality, specifically tenderness.
The standard Warner-Bratzler shear force test may measure only a component of the factors affecting tenderness, while elastography may be more sensitive to specific influences such as the effects of collagen.
Ultimately, elastography may be able to assess marbling in chilled carcasses, Ophir suggests.
"After chilling, the fatty areas are harder than lean muscle," he points out. "Ideally, we could evaluate marbling using elastography to differentiate between fat and lean. Eventually, we may be able to apply the technology to live animals."
TenderTecª
Also in the quality area, an innovative carcass probe developed by the Meat Research Corporation of Sydney, Australia, holds the promise of what some researchers hope will be fairly reliable data on beef tenderness, a key element of eating quality. (The unit is under study in the United States, but results to date are preliminary and unavailable for publication).
The TenderTec system uses a probe that is "twisted" into the muscle tissue at three varying depths. An instrument measures the amount of force needed to rotate the probe, and readings are taken at each probe depth as a measure of the beef's elasticity.
"The TenderTec probe relies on a special 'head,' which the inventor claims is crucial to its operation," says Tony Gordon, a scientist with Meat Research Group who has been working extensively with TenderTec. "The device has possibilities, but right now we're still trying to get the unit up to a level of commercial reliability."
Gordon says that TenderTec may have application in Australia for detecting the "odd carcass" that is extremely tough.
"We have a lot more variability in our cattle here," he points out. "Whether it would be as useful in the states [for that purpose], I don't know."
While TenderTec is still "under assessment" in Australia, Gordon stops well short of endorsing the unit. "It's still unproven technology in my mind," he cautions.
He also points out the limitations inherent in using any single instrument to assess beef quality.
"You can't come up with a true quality measurement using only a device like TenderTec," he says. "For example, our research has found that there's little correlation between consumer sensory data and objectively measured strip loin tenderness."
On the subject of beef quality in general, Gordon concludes-no pun intended-"It's pretty tough to predict tenderness."
Swatland's probe
Such difficulties haven't stopped development of an innovative carcass probe by Canadian researcher Howard Swatland, professor of animal and poultry science at the University of Guelph in Ontario. His system attempts to measure connective tissue in beef animals as a way of predicting beef palatability.
Based on a Danish device used to measure fat percentage, Swatland's customized probe uses a single optical fiber connected directly to an optical box behind a computer. As the probe is inserted into a beef carcass, connective tissue "fluoresces" when hit with a UV light beam, sending light back up the fiber.
The fluorescent light has a different wavelength and can be measured using software that plots a graph detailing the connective tissue "matrix" of each carcass. Swatland's research suggests a close relationship between the presence of connective tissue and palatability.
But the inventor himself cautions that the issue is complex.
"There are many sources of toughness in beef," Swatland points out. "But if you control for the key factors-electrical stimulation, cold shortening and improper aging-then this system delivers up to an 80 percent correlation between probe data and ratings from a consumer taste panel evaluating beef tenderness."
Although commercial viability remains a few years away, Swatland says potential applications include using probe data to select the top 10 percent of carcasses to be vacuum packed and aged for the HRI (hotel, restaurants and institutions) segment.
"The probe will identify beef that's not full of connective tissue-and I can guarantee that beef will be good quality," he says.
Other uses include selecting acceptably palatable beef from lower quality sides-a process being explored in Europe, where much of the beef comes from older dairy animals-and identifying connective tissue structure patterns that could objectively estimate the age of live animals at slaughter.
"Right now, estimating age is a subjective decision graders must make," Swatland notes. "It would greatly benefit the industry if that process could be standardized."
Also under development is a miniaturized probe the size of a bovine veterinarian needle that Swatland says could potentially assist seedstock operators in identifying "rogue" animals that carry the genetics to produce excessive connective tissue in their progeny. "It's preferable to pull these animals out than try to select for traits affecting connective tissue in general," he says.
Overall, Swatland freely admits his probe is very much in the developmental stage. But he envisions near-term applications ranging from assessing live animals to detecting cold-shortened beef in the cooler.
The future
Overall, the outlook among researchers and packers alike is equally sanguine. The beef grading instrumentation now being tested, and in some cases utilized, offers the promise of not only faster, more accurate information on yield and quality, but more importantly, at least a partial solution to the economic challenge of making beef production and processing more economically viable.
"To remain competitive, we need to do a better job of assessing beef quality," says John Stowell, associate director of beef improvement and education for the National Cattlemen's Association.
"We can't rely on [human] graders who have to check more than 300 carcasses an hour. Don't get me wrong-they do an excellent job, but we could do it better with technology," he adds.
Doing a better job translates into reaping a greater reward, of course, and that remains the bottom line for investing in these high-tech systems.
Making the Grade
As technology pushes toward automated beef grading, what will happen to the thousands of USDA graders?
by Dan Murphy, contributing editor
Perhaps no other task in the beef industry is so crucial, yet so subjective as grading.
A virtual army of USDA personnel takes on the daily task of grading more than 125,000 cattle in the nation's beef packinghouses, and the results significantly affect the economic value to the entire production and processing chain.
A quick primer on beef grading: Designed to separate highly variable cattle in generalized groupings, USDA yield grades range from YG1 to YG5. The lower the number, the higher the yield. A complicated formula for calculating yield grades combines adjusted fat thickness, kidney-pelvic-heart fat percentage, rib eye surface area, and hot carcass weight.
That's the procedure on paper. In reality, the grade must be assigned by sight in a matter of seconds as cattle are railed past the inspector. Hence the need for technology such as video imaging that can automate and standardize grade assignments.
New technologies
"Some of the new technologies, such as video image analysis (VIA), may eventually take over routine yield grading," acknowledges Jim Wise, meat marketing specialist in USDA's Agricultural Marketing Service. "Conceivably, VIA could calculate the majority of carcass grades, freeing the inspector to handle the exceptions where fat variations or contour defects would affect the reading."
Others are more blunt. "VIA already does a very good job at yield grading cattle," says John Stowell, associate director of beef improvement and education for the National Cattlemen's Association. "The technology is as good as USDA [grading] and has less variability."
Although VIA is still a ways from replacing human graders, it may have a short-term benefit. Wise says a sophisticated VIA system may spur efforts to split the yield grades. USDA specifications have already been developed for use of YG 2.0, 2.5, 3.0, 3.5 and so on, but no one has filed for approval to use them yet.
"It costs the packers additional time and additional training," he admits. "It's a lot more difficult to judge grades when they are divided up, especially at today's line speeds."
Split grading will benefit both packers and producers, Wise argues. For example, he says packers could use VIA to better segment cattle destined for one-fourth-inch trim programs. Yield data could also be sent upstream to producers, allowing them to fine-tune breeding and management programs.
But will VIA lead to loss of jobs among the current cadre of USDA graders? Wise says no-at least not anytime soon.
"Near term, VIA won't replace anybody," he says. "Graders may need additional training, but they will still be needed to perform quality grading, even with the assistance of a video system. Eventually, USDA may perform more of a monitoring function. We might even end up as a third party to packer-run systems someday. This technology has an awful lot of promise."