8fbd035081bd09934004bfc61d79c31c5d5d9ee4

Pet tech boom: biometrics and AI gain favour among cat and dog owners

From biometric ID to predictive diagnostics, AI is helping pet owners and veterinarians improve the quality of life of our burgeoning population of dogs and cats.

Anyone who’s noted the increased numbers of dog walkers in their neighbourhood will know that the Covid-19 crisis unleashed a pet ownership boom. Millions acquired four-footed companions to help them through the stresses of home isolation.

The People’s Dispensary for Sick Animals found that 36% of all current UK owners of dogs, cats and rabbits acquired their pets after March 2020, indicating that 4.1 million dogs and 3.7 million cats have been homed since the onset of the pandemic. 

Pet population rises synced with other digital dependencies heightened by lockdown. Our use of connected technology helped to manage routine but essential aspects of daily life, from grocery delivery to virtual medical check-ups. And as more pet food was added to online shopping baskets, it became clear that a new generation of pet parents looked to tech for care and support for new furry family members.

The burgeoning pet tech market has, in turn, fuelled the need to leverage on-demand cloud-based artificial intelligence (AI) resources to power pet care platforms for use by owners and veterinarians alike.

The market is projected to be worth $22bn by 2028, with a very healthy compound annual growth rate of 23.2% over the forecast period. Increasing demand for remote monitoring and tracking of pets, plus technology for securing pets against theft and accidental loss, is driving growth in the global pet tech market.

The trend is underwritten by pet owners’ continued propensity to see their pets as investments in personal and familial wellbeing. As a result, they are now willing to spend more money on their pets’ medical care and fitness, as well as aids to their pets’ mental health and bodily security. But all this additional pet parenting adds to already busy domestic schedules. Owners want ways to automate the tasks of pet management through technology, including smart collars that track health and location, pet observation through webcams, smart pet feeders and robotic fitness toys.

Key to this emergent market are, of course, the capabilities of smartphone functionality. Smart endpoint devices are serving as remote diagnostic data capture tools that can monitor symptoms and vital signs to check for healthcare issues. Smartphones are also serving as frontline health scanning devices across a range of different pet care applications. 

Seoul-based Petnow’s dog biometric ID app, for example, uses AI software that automatically scans the unique canine identifiers usually known as ‘nose prints’ – the wrinkle patterns on the outer skin tissue around the nostrils (although the software also cross-checks this against a scan of the animal’s face). Petnow uses automated extraction and recognition technology, and promises real-time object detection and classification of dogs following a photographic scan.

“When a dog owner runs the Petnow app, three types of AI and machine learning will assist users in nose print registration,” explains Dr Jesse Joonho Lim, CEO at Petnow. “If the quality is insufficient, the software automatically retries the acquisition process to obtain higher quality images.”

By registering a dog’s profile with the nose print, the owner is provided with an additional method to be notified in case the dog goes missing and needs to be positively identified to prove ownership. Petnow has announced plans to add cat biometric recognition to the app before the end of 2023.

Spot the dog’s symptoms

TTcare, from AI solutions provider AI FOR PET, is designed to use photographs of cats and dogs for healthcare purposes. Launched in South Korea in 2021, the app is configured to detect early indications of diseases and health issues in dogs and cats. Following its launch in the US and Germany in 2022, the TTcare platform now has about 140,000 registered users.

After capturing a digital image of the pet, TTcare’s AI software analyses the image and informs the animal’s owner of potential eye, epidermal or joint-related conditions and diseases. TTcare’s AI is trained to recognise physical manifestations of potential problems to foreshorten the time to initial or preventative treatment, says Jay Lim, marketing team leader at AI FOR PET.

“The images taken by TTcare users are used in AI relearning. As data they are labelled by professors at Seoul National University and Konkuk University’s veterinary school,” Lim says. “We also gather image data from animal hospitals and dog and cat shelters to improve the [teaching and] accuracy of the AI, as well as to develop new features, such as [the monitoring of] joints and teeth.”

Lim adds: “To be exact, TTcare does not diagnose medical conditions, but informs users of evidence of disease-related abnormalities. For dogs, TTcare’s AI can analyse and give information on 11 signs for the eye and four signs for the skin that may potentially lead to diseases that can be detected externally. For cats, TTcare can analyse four abnormal signs related to eye diseases.”

The AI used by TTcare has its limitations. It cannot be trained to recognise abnormal signs related to ailments that are specific to certain breeds of dogs or cats, for example. “We do, however, provide information on diseases that occur frequently, or that owners should be more careful about, based on the breeds of pets entered into the app by the user,” says Lim.

TTcare is now being developed to use full-motion video capture as the basis of AI-powered symptom recognition for canines. Lim says: “TTcare’s joint program detects the abnormal movements of the legs associated with certain diseases [using] a video capture of the movement of the dog. We also plan to provide functions that can analyse the teeth, and that can determine [health] emergencies from checking stool, vomit and heart rate.” 

AI FOR PET developers are also investigating the potential to adapt TTcare for use with other animal species or agricultural animals. “It’s not going to be easy,” Lim concludes, “but we would like to develop TTcare into a service that can manage the health of animals such as horses, cows and pigs.”

Relaxed, not rattled

Joii Pet Care is an app from UK-headquartered veterinary technology company Vet-AI. It was launched in 2019 to offer pet owners remote consultations with vets and veterinary nurses in real time via smartphones. More than 330,000 pets are now registered on the app, according to Vet-AI.

The company says that in some cases an online consultation is preferable for vets because it enables them to view a pet in its own domestic environment. The animal is more likely to be relaxed than rattled by the surroundings of a veterinary clinic.

“Joii Pet Care has analysed that some 70% of pet healthcare issues are resolved without the need to go into a vet practice,” says Dr Samantha Webster, director of clinical operations at Vet-AI. “The app’s symptom checker tool can identify more than 160 common conditions quickly – informing and supporting the pet owner on next steps.”

Consultations are not restricted to physical or metabolic ailments. Increasingly, Joii Pet Care’s AI is being used to track behavioural problems in dogs, which have been on the rise since 2020. Behavioural problems remain one of the top reasons for dogs to be surrendered to rescues and can be a cause of stress for their owners.

“The most common behavioural issue seen by Joii Pet Care is that of separation anxiety,” Webster says. “There’s a growing feeling in the behaviour community that, in some cases, it can be beneficial to treat behaviour remotely.

“By not entering the pets’ home, the behaviourist is able [to use the Joii Pet Care app] to see the normal day-to-day interactions of pets and owners without influencing them, and therefore get a better understanding of how to help with any problems.”

Combating cat killers

The UK population of domestic cats is estimated to be at approximately 11 million – and our feline friends are living longer. The average feline lifespan has gone from seven years in the 1980s to around 14 years in 2015. 

Chronic kidney disease (CKD) remains a major cause of death for cats. Conventional detection methods find it only after significant kidney damage has occurred, complicating intervention and treatment. With appropriate medical management, however, cats with CKD can survive for a significant amount of time with a reasonably good quality of life.

RenalTech is a veterinary diagnostic tool that uses AI and machine learning to predict CKD with greater than 95% accuracy – predicting whether a cat will develop CKD up to two years faster than traditional clinical diagnosis. RenalTech relies on six common feline health parameters – blood urea nitrogen, creatinine, urine specific gravity, urine protein, urine pH and white blood cell count – and approximate age. This is all taken from a database during routine wellness visits to deliver a RenalTech status.

The tool was developed by analysing the anonymised medical data of more than 150,000 cats and 750,000 patient visits over 20 years to a Vancouver pet hospital. The ability to identify whether a cat will likely develop CKD enables vets to confer with owners of at-risk animals on proactive, tailored healthcare plans – such as renal therapeutic diets and CKD-related medications. 

As has proved the case with human patients, AI-enabled pre-emptive screening will, going forward, lower the number of surgery appointments and improve quality-of-life expectations. However, stark differences exist between the medical treatment of people and that of their companion animals. Because of the nature of the client-patient-practitioner relationship, the use of AI in veterinary practice is already giving rise to debates around wider ethical issues – debates that doting owners will increasingly want to be part of.

Pet pronouncements…

Cat communicator app gives felines a say

Akvelon’s MeowTalk is a cat ‘translation’ app that leverages machine learning to help owners understand what messages their pets are trying to convey in their mews and purrs. 

 “Every cat has a unique vocabulary that they consistently use to communicate with their owners,” says Javier Sanchez, chief product officer at Akvelon. “Each cat produces distinct meows for intents such as ‘Feed me’, ‘Let me out’, ‘I am happy’ and, perhaps most importantly, ‘I am in pain.’ 

“The app recognises nine built-in ‘intents’ that it can detect right out of the box and are general to all cats. We are using advanced machine learning and data science to go beyond just guessing or picking up on tones.”

Users can also train MeowTalk’s machine-learning technology to translate additional meows that are specific to a given cat’s vocabulary so that the app can fully understand and translate most of those intents.

“Akvelon has leveraged a stack of Google Cloud technologies for MeowTalk’s back-end and data science pieces,” Sanchez says. “We’re going beyond just informed guesswork or picking up on tones. We’re matching and detecting specific cat meows similar to how smart speakers pick up on human speech.”

The app has already acquired more than one million users worldwide. Akvelon technologists are also developing a smart collar that can recognise cats’ meows and instantly provide an audible translation, giving a wearing cat “an expressive voice”.

Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.

Related Posts

Recommended Articles