CLOSE
Original image

Inside the Surprisingly Delicious World of Cat Food Taste Testing

Original image

Cat food is serious business. Taking underutilized and low-value raw materials like animal byproducts and turning them into high-value foods can be, not surprisingly, very lucrative. Along with other pet foods, cat food makes up a hefty portion of the international prepared foods market.

The crowns of the cat food kings are heavy, though. Their products have to be palatable and nutritious for cats, as well as convenient and economic for the owner. Accomplishing that first part isn’t easy when many of their customers are sensitive to even subtle flavor differences, very picky about their food, and can’t even verbalize what they think of the product.

Behavioral studies on cats can give the food producers a little feedback, but they’re often limited to very simple acceptance and preference tests that are time-consuming, complicated by variations among different individual cats and, in the end, not very data-rich. Facing these limitations in gauging the likes and dislikes of cat food’s end-users, brilliant minds in industry and academia put forth the idea of nixing four-legged taste testers in favor of two-legged ones.

The Truth About Cats and Humans

Yes, there are differences in cats’ and humans’ physiological and perceptual systems, but there are also some similarities, as well as experimental evidence that human sensory data could be useful in cat food formulation. Human taste tests could be done, sure – Simon Allison, a senior food technologist at UK retailer Marks & Spenser, has admitted that, by his own choice, he tastes all of the company’s products – but how? And would they do any more good than cat taste tests?

In 2007, Dr. Gary Pickering, currently a Professor of Biological Sciences and Psychology/Wine Science at Brock University in St. Catharines, Ontario, set out to develop a methodology for using human taste panels to assess canned cat food. The panel of taste testers was drawn from the staff and student population of Charles Sturt University-Riverina in Australia, where Pickering taught at the time, and screened with a battery of tasting exercises. In the last exercise, Pickering got down to the nitty gritty and brought out the cat food.

Let's Hear From Our Judges

The prospective panelists tasted three different canned cat foods and rated their “hedonic impression” (whether they liked or disliked it) on a 9-point scale. This helped to weed out people who were really grossed out over or hated eating the cat food and, hence, might have reduced motivation, concentration or reliability in the study. About 1/3 of the prospective panelists opted not to continue with the screening process, with dislike of the cat food being most common reason for withdrawing. (Shock!)

The final panel – consisting of 11 who apparently didn’t completely hate the act of eating cat food – rated samples of cat food meat chunks, gravies/gels and meat-gravy mixes over the course of six tasting sessions. They were first asked to describe the samples’ flavors and textures using a descriptor generation form provided by Pickering, resulting in a list of 119 flavor and 25 texture descriptors. That list was pared down to 18 flavor descriptors: sweet, sour, tuna, herbal, spicy, soy, salty, cereal, caramel, chicken, methionine, vegetable, offal-like, meaty, burnt, prawn, rancid and bitter. There were also four texture dimensions: hardness, chewiness, grittiness and viscosity. The panel then scored a range of cat food products for intensity of each of the flavors on the list and for “hedonic impression.”

These tastings, and the flavor attributes and intensity ratings they generated, allow for flavor profiles to be developed for individual cat food products. The finer details of the usefulness and limits of human taste testing of cat food still need to be worked out—for example, cats don’t have a sweet taste receptor, so the human detection and rating of that taste doesn’t do anyone any good. But the combination of these flavor profiles and the cat acceptance/preference studies already in use could enable faster, more economical ways of optimizing cat food flavor and texture and predicting the effects that any changes to the food might have on picky kitties.

Mikey Likes It (Slightly)!

While that practical application of the results is all well and good (go science!), the real take-away for me is this: Canned cat food apparently doesn’t taste as gross as it looks, smells and feels, and it’s for the strangest reasons. The average (mean) of all the panelists’ hedonic scores was 4.97 on the study’s 9-point scale, right between “neither like nor dislike” and “like slightly.” Not bad! Even more surprising is that positive, or “like,” scores were positively correlated with rancid, offal-like, burnt and bitter flavors, but negatively correlated with tuna and herbal flavors.

Reference: “Optimizing the sensory characteristics and acceptance of canned cat food: use of a human taste panel.” Journal of Animal Physiology and Animal Nutrition, Volume 93, Number 1, February 2009, pp. 52-60(9). Published online: February 2008. DOI: 10.1111/j.1439-0396.2007.00778.x

Original image
iStock // Ekaterina Minaeva
technology
arrow
Man Buys Two Metric Tons of LEGO Bricks; Sorts Them Via Machine Learning
May 21, 2017
Original image
iStock // Ekaterina Minaeva

Jacques Mattheij made a small, but awesome, mistake. He went on eBay one evening and bid on a bunch of bulk LEGO brick auctions, then went to sleep. Upon waking, he discovered that he was the high bidder on many, and was now the proud owner of two tons of LEGO bricks. (This is about 4400 pounds.) He wrote, "[L]esson 1: if you win almost all bids you are bidding too high."

Mattheij had noticed that bulk, unsorted bricks sell for something like €10/kilogram, whereas sets are roughly €40/kg and rare parts go for up to €100/kg. Much of the value of the bricks is in their sorting. If he could reduce the entropy of these bins of unsorted bricks, he could make a tidy profit. While many people do this work by hand, the problem is enormous—just the kind of challenge for a computer. Mattheij writes:

There are 38000+ shapes and there are 100+ possible shades of color (you can roughly tell how old someone is by asking them what lego colors they remember from their youth).

In the following months, Mattheij built a proof-of-concept sorting system using, of course, LEGO. He broke the problem down into a series of sub-problems (including "feeding LEGO reliably from a hopper is surprisingly hard," one of those facts of nature that will stymie even the best system design). After tinkering with the prototype at length, he expanded the system to a surprisingly complex system of conveyer belts (powered by a home treadmill), various pieces of cabinetry, and "copious quantities of crazy glue."

Here's a video showing the current system running at low speed:

The key part of the system was running the bricks past a camera paired with a computer running a neural net-based image classifier. That allows the computer (when sufficiently trained on brick images) to recognize bricks and thus categorize them by color, shape, or other parameters. Remember that as bricks pass by, they can be in any orientation, can be dirty, can even be stuck to other pieces. So having a flexible software system is key to recognizing—in a fraction of a second—what a given brick is, in order to sort it out. When a match is found, a jet of compressed air pops the piece off the conveyer belt and into a waiting bin.

After much experimentation, Mattheij rewrote the software (several times in fact) to accomplish a variety of basic tasks. At its core, the system takes images from a webcam and feeds them to a neural network to do the classification. Of course, the neural net needs to be "trained" by showing it lots of images, and telling it what those images represent. Mattheij's breakthrough was allowing the machine to effectively train itself, with guidance: Running pieces through allows the system to take its own photos, make a guess, and build on that guess. As long as Mattheij corrects the incorrect guesses, he ends up with a decent (and self-reinforcing) corpus of training data. As the machine continues running, it can rack up more training, allowing it to recognize a broad variety of pieces on the fly.

Here's another video, focusing on how the pieces move on conveyer belts (running at slow speed so puny humans can follow). You can also see the air jets in action:

In an email interview, Mattheij told Mental Floss that the system currently sorts LEGO bricks into more than 50 categories. It can also be run in a color-sorting mode to bin the parts across 12 color groups. (Thus at present you'd likely do a two-pass sort on the bricks: once for shape, then a separate pass for color.) He continues to refine the system, with a focus on making its recognition abilities faster. At some point down the line, he plans to make the software portion open source. You're on your own as far as building conveyer belts, bins, and so forth.

Check out Mattheij's writeup in two parts for more information. It starts with an overview of the story, followed up with a deep dive on the software. He's also tweeting about the project (among other things). And if you look around a bit, you'll find bulk LEGO brick auctions online—it's definitely a thing!

Original image
Nick Briggs/Comic Relief
entertainment
arrow
What Happened to Jamie and Aurelia From Love Actually?
May 26, 2017
Original image
Nick Briggs/Comic Relief

Fans of the romantic-comedy Love Actually recently got a bonus reunion in the form of Red Nose Day Actually, a short charity special that gave audiences a peek at where their favorite characters ended up almost 15 years later.

One of the most improbable pairings from the original film was between Jamie (Colin Firth) and Aurelia (Lúcia Moniz), who fell in love despite almost no shared vocabulary. Jamie is English, and Aurelia is Portuguese, and they know just enough of each other’s native tongues for Jamie to propose and Aurelia to accept.

A decade and a half on, they have both improved their knowledge of each other’s languages—if not perfectly, in Jamie’s case. But apparently, their love is much stronger than his grasp on Portuguese grammar, because they’ve got three bilingual kids and another on the way. (And still enjoy having important romantic moments in the car.)

In 2015, Love Actually script editor Emma Freud revealed via Twitter what happened between Karen and Harry (Emma Thompson and Alan Rickman, who passed away last year). Most of the other couples get happy endings in the short—even if Hugh Grant's character hasn't gotten any better at dancing.

[h/t TV Guide]

SECTIONS
BIG QUESTIONS
BIG QUESTIONS
WEATHER WATCH
BE THE CHANGE
JOB SECRETS
QUIZZES
WORLD WAR 1
SMART SHOPPING
STONES, BONES, & WRECKS
#TBT
THE PRESIDENTS
WORDS
RETROBITUARIES