How biometric wearables could transform well-being for dogs
There is increased demand for monitoring pet health. But how useful and reliable is the data collected by ‘smart’ devices?
Humans have benefitted from on-the-wrist heart- rate tracking for years, and the dog equivalent seems on the verge of going mainstream this year. While a number of products offering doggy biometrics via a collar or harness have been promised in recent years, 2023 is shaping up to be the tipping point as multiple new products are entering the market.
Early warning of health issues
French consumer electronics company Invoxia is about to launch its Smart Dog Collar priced at €169 ($182), with a compulsory €8 ($8.60) per month subscription. It not only promises up to 15 days of GPS tracking on a single charge, but also continuous tracking of the dog’s cardio-pulmonary system via the company’s Heartprint technology. This provides early warnings of health issues via the connected app.
“Dogs can’t express themselves to say something is wrong,” says Invoxia CEO Amélie Caudron to PETS International. “This means that oftentimes we wait until it’s too late or until it’s a very complicated situation before taking them to the vet.”
For example, Caudron claims that the collar can detect tell-tale symptoms of Mitral valve disease –a heart condition that’s very common in King Charles Spaniels – allowing vets to intervene early, which potentially adds months to a dog’s life. She cites a recent study suggesting that the early diagnosis of congestive heart failure could give the pet an extra 10 to 15 months of life. “Which is huge given that a dog only lives 10 to 15 years,” she states.
Use of data
The company hopes that when enough dogs worldwide are wearing the collar, it will be possible to build a database of canine biometric data of unprecedented scale. After all, clinical results aren’t necessarily representative due to the extra stress dogs are under in a research setting. And Caudron believes the benefits go beyond refining existing knowledge.
“From this database, we’ll be able to find new markers of pathologies, to find correlations,” she continues. “The more dogs that wear our product, the more correlations we can find and the more diagnostically powerful the data becomes.”
Tracking the animal’s heartbeat
While human wearables use optical heart-rate sensors to detect heart rate, such an approach won’t work with most canine subjects due to the coarse hair disrupting measurements. Radar is an attempt to sidestep this problem. The technology involved in Invoxia’s solution is miniaturized radar sensors combined with edge artificial intelligence (AI) to track biometrics.
Taiwan’s Industrial Technology Research Institute (ITRI) has a similar system in place with its iPetWeaR collar technology. “We use the small radar sensing technology that can be put on the animal’s neck to detect heart rate and respiration,” explains Director Dr. Hong-Dun Lin.
“You can use the app to detect most vital signals without the removal of hair from the animals,” Lin continues. The company also has plans to move beyond furry patients onto ‘husbandry applications’ for animals like cows. A number of Taiwanese companies are already interested in licensing the technology this year, and availability in Europe is anticipated within a year. The cost of each device will be less than €93 ($100) and will likely be supplemented with a subscription for data analysis.
Skepticism
Not everyone is convinced that the technology is ready yet. Dr. Firat Güder, an expert on intelligent interfaces at Imperial College London’s Department of Bioengineering, is one such skeptic, having researched the area himself for years.
“There are a few companies that say they’re going to release products that can do this, but most of these companies never publish their data,” he explains. “It’s my view that most of these collars do not really work well, especially at the price ranges where people are claiming to make these things work.”
In his own research, Güder has run into 3 major problems: the diverse range of dog breeds, the excitable temperament of dogs and, of course, the hair. “Before an operation, people with minor chest hair will be shaved to put electrodes on – and human hair is nothing compared to a dog with 10-centimeter-long hair all over its body.” He’s still not convinced about the ability of radar technology to overcome the hair problem: “Even that would kind of struggle to work reliably all of the time.”
Is it viable?
There is also a variability problem, he explains: “Small dogs have different heart rates than bigger dogs. And then even bigger dogs, depending on the breed, will have different physiological manifestations of their metabolism.” The good news? None of this is insurmountable, according to Güder. But so far, he has seen no evidence to suggest that reliable retail solutions are imminent, so he is concerned that companies are pushing human solutions on a fundamentally incompatible species.
Instead, he advocates a bottom-up solution led by engineers and scientists, but that requires serious investment in terms of both people and time. “We’ve been working on this for around 9 years now, and we’re only just at a place where we feel comfortable that what we know is enough to be able to solve problems,” Güder says. His newly formed company is aiming to commercialize stretchable sensors for more reliable body contact. He claims that “pretty good prototypes” are expected within the next 2 years.
Despite this, the companies with more ambitious release dates in mind seem pretty confident in their data. Invoxia claims 98% accuracy with support for 90% of the top dog breeds in the US and Europe (albeit based on an unrevealed sample size).
But the real test will come when the first smart collars adorn the necks of pets around the world. If the data isn’t reliable enough for vets to take seriously, then owners may quickly cancel their subscriptions, leaving a seemingly promising market sector dead on arrival.