Introduction

This project aims to address the phenomenon of Singaporean dog owners having insufficient information on the nutritional needs of their pets (Khoo, 2022). There is a tendency for dog owners to overfeed their dogs as a consequence, resulting in the abnormally high prominence of obese dogs in Singapore. This phenomenon is not unique to Singapore, where 80% of dog owners do not measure the amount of food fed to their pet, blatantly ignoring the feeding guides provided on food packaging (Royal Canin Weight Management Congress, 2018).  In Singapore the overfeeding situation is dire with an alarming 40% to 50% of dogs being categorised as overweight or obese, and these numbers are still increasing exponentially (The Animal Clinic, 2020). Our bioelectrical impedance analysis (BIA) scale and automated pet feeder innovation aims to combat these problems by providing dog owners with more accessible information about their dog’s health and streamline the feeding process by having the desired amount of food measured out automatically and fed to the dog.

Problem Situation

Significance and impacts
Undoubtedly, obesity is one of the most common nutritional disorders seen in canines (German, 2006) and contributes significantly to a myriad of health problems in dogs, such as increasing their risk of developing chronic diseases and even mortality. Overweight dogs are more prone to a never-ending list of possible diseases, such as cancer, diabetes mellitus, hypertension and heart diseases (Weir, n.d.). Obesity reduces a dog’s life expectancy by an average of 6-12 months, where a large lifetime study of Labrador retrievers revealed that just being moderately overweight reduces a dog’s life expectancy by 2 years (Weir, n.d.).

Target group analysis
Through this project, we aim to help reduce canine obesity, especially in Singapore where it is one of the most common issues faced by dogs. We wanted to come up with a solution that was not only practical for users to measure body composition but was also user friendly so that they can easily track the changes in body composition for dogs. After much research, we came upon bioelectrical impedance analysis (BIA) as a good method that fulfils the criteria we had set out with. 

BIA measures body composition by measuring the body’s impedance when a small, non-lethal current is run through the body. BIA not only directly measures body composition but it is also not invasive and does not require sedation. While BIA is not as accurate as DEXA, the technology behind it has been refined through years of research, leading to more and more accurate machines that are now comparable to DEXA (Arreesrisom et al., 2022). BIA is also now widely available for human use (Catapano et al., 2023), further cementing it as a credible approach.

Our product will have 2 different parts: A BIA weighing scale custom built for dogs as well as an automatic dog feeder.

The BIA weighing scale will allow owners to weigh their dogs as well as measure their body composition. BIA is able to measure body composition by measuring impedance. The measured impedance is directly related to the total body water volume. As muscles store most of the body’s water, the more muscles a dog has, the lesser impedance will be measured. Combining the weight of the dog and these values, we can find out how much fat mass a dog has. This allows owners to easily keep track of their dog’s fat mass without making multiple appointments to the vet. 

As BIA scales for humans are already commercially available, we plan to adapt components of a human BIA scale for dogs. For humans, the electrodes that allow a current to run through the body are placed on the weighing scale itself. While that is practical for humans, where the foot length has minimal differences, that is not the case for dogs. Different dogs will have drastically different distances between their legs, making a human BIA scale unsuitable for use for dogs. Hence, we intend to create a BIA scale where the electrodes are not fixed to the scale but can instead be attached to the dog itself. The electrodes will be similar to electrocardiogram (ECG) electrodes which are sticky and made of gel, allowing us to securely attach the electrodes to dogs of all sizes and shapes. 

This leads up to our next part: the automatic dog feeder. With the body composition of the dog, we will put this data through an algorithm that will calculate how much kibbles the dog can consume and then dispense it. This algorithm takes into account the breed, age and nutritional content of the kibbles dispensed when computing the appropriate amount for the dog. 

In this way, we do not only come up with a device that measures the body composition of a dog, it also helps owners to regulate the diet of their dogs, reducing the situation of overfeeding.

Root causes
Canine obesity is caused by many different factors but it typically stems from the following two factors:

1. Owners are unaware that their dogs are obese (Ronja & Kölle, 2021) and hence make no lifestyle changes to curb this problem.

Currently, obesity in dogs is diagnosed mainly by visual examination, which is known as body condition scoring (BCS). BCS is widely adopted by veterinarians as it is an easy and simple way to gauge a dog’s body composition, offering an assessment that correlates well with percent body fat (Mawby et al, 2004, Shoveller et al, 2014). However, it is semi-qualitative as it does not directly measure body composition and its accuracy is largely influenced by the assessor’s experience and knowledge (Boswald et al., 2019). Hence, owners can only determine if their dogs are obese when they go for vet appointments and it may be too late by then.

Another method to determine body composition is dual-enery X-ray absorptiometry (DEXA) (Zanghi et al., 2013). While it is very accurate, it requires specific equipment and sedation. This makes it impractical and unfeasible for vets to measure body composition. The machine used for measuring DEXA in animals is niche and not available in Singapore thus far, as it is specifically adapted from humans for animals and requires intensive calibration to produce accurate DEXA results for animals. In recent years, only a machine for small animal DEXA (e.g. rats) has become more accessible, but one for larger animals is still not available.

The last method available for measuring body composition is by determination of ash weight and bone histomorphometry. However, these methods of analysis require bone samples that can only be acquired after the animal is dead thus it is simply not feasible. Hence, this shows that dog owners currently have no means of obtaining the body composition of their pets. Thus our product is very necessary as a tool in helping pet owners take charge of their pet’s health. 

2. Owners are uneducated on the right amount of food to feed their dogs, often overfeeding them (Grota, 2022).

Many owners are not well-educated when it comes to the diets of their furry companions. A major reason for this is because of how difficult it can be to determine how much as well as what a dog should eat. There are so many different breeds of dogs that require different diets (Alegría-Morán, R. A., S. A. Guzmán-Pino, J. I. Egaña, C. Muñoz, and J. Figueroa, 2019) and this is not even considering crossbreeds! Combining this with a dog’s metabolism, lifestyle and health conditions, it is clear to see how complicated coming up with a suitable diet is (Gilbert, n.d.).

It can also be hard to tell how varying different components of a diet can affect dogs. From a 2003 study, it was proven that even just slightly increasing the fat content of a dog’s diet can double the fat mass in a dog even as its body weight remains relatively stable (Kim et al, 2003). Different brands of kibbles also have different nutritional content. While there are instructions to guide owners on how much kibbles can be eaten a day, different dogs of differing ages would all have different nutritional needs. Afterall, a chihuahua definitely doesn’t need as many calories as a german shepherd. Not only that, as a dog grows older, they require less nutrients and may require food with a different moisture content (Kim, K. H., Seo, K., Cho, H.-W., Jeon, J.-H., Kim, C. H., Jung, J., & Chun, J. L. ,2021). With all these considerations that owners need to take into account, it is no surprise that some owners do not have enough knowledge to devise a suitable diet. Hence, owners are prone to overfeeding their dogs.

Working Principle

Our innovation is very relevant as in tackling obesity, preventive measures precedes reactive measures since obesity is much harder to address once established (Zoran, 2010). Furthermore, quality of life based on various factors such as vitality, chronic pain that is often improved with successful weight loss has been positively correlated with changes in body fat mass (German, 2016). Having the ability to monitor and track the fat mass of a dog brings about substantial benefits to a dog’s health and helps reduce the dog’s obesity. 

BIA scale

Objectives:
Our BIA scale provides easily accessible body composition details of a dog at the press of a button and most importantly, identifies the amount of excess body fat in a dog. This helps owners take charge of their dog’s health, be it to identify whether their dogs are obese, track their dog’s weight-loss journey or be alerted to warning signs that their dogs are close to being obese (preventive measure). In a bid to enhance user experience with our device, we will also be developing an app which allows owners to track past and present records easily.

Details:
At its core, bioelectric impedance works on the basis that different tissues (namely fat, muscle and water content) in the body have different electrical conductivities. This allows for an estimation of body composition by measuring the impedance of a body towards a small electrical current since electric current passes through the body at a differential rate depending on body composition (Dehghan & Merchant, 2008). Results obtained from our BIA scale are validated against BCS.

We decided to go with multi-frequency bioelectrical impedance analysis (MFBIA) as we found that when compared to single bioelectrical impedance analysis (SFBIA), MFBIA is more accurate (Holmes et al., 2022).

Limitations:

• Hydration Levels: Body fat scales assume that the body’s hydration levels are relatively constant. However, dehydration can skew readings by reducing the body’s water content and increasing impedance, leading to an overestimating body fat percentage.
• Body Position: Body fat distribution can vary depending on body position. For example, standing on the scale with different foot placements can affect the readings obtained
• Different breeds: Every dog breeds have different body composition which calls for a need to adjust the algorithm to account for different dog breeds.

Pet feeder

Objectives:
To cater to the individual nutritional needs of the dog, according to the results of the BIA analysis. This helps the owner to accurately determine the amount of food his/her dog requires, acting as a measure against canine obesity. 

Details:
Owners will measure the body composition of their dog using the BIA scale, as well as input the nutritional content (mainly calories) of the brand of kibbles the dog is fed and the breed of the dog. The system will process these inputs, using algorithms and databases that we will be coding, and dispense the accurate amount of food his/her dog needs. Similarly, we will integrate the algorithm into our app in a bid to enhance user interface. The customised amount of food to be fed will be automatically calculated and dispensed by pressing a button on the app.

As a rough guide, a dog’s daily caloric need, resting energy requirements (RER), can be calculated using the following formula: 30 x [dog’s weight in kilograms] + 70. For obese dogs, we will first recommend dog owners to feed their dog according to the RER. However, if there is no change in the dog’s weight after 2 to 3 months, we will advise dog owners to cut down the calories further by 10%, a safe and controlled amount which will help to further facilitate weight loss (Cox, 2023).

Limitation:

• Daily calories intake is highly dependent on breed: Different breeds differ in sizes and activeness, hence daily calories intake differs as well. Given the numerous breeds of the dogs, it is difficult to account for all breeds. However, this will be a future outlook of our project, with continuous updates on the database to include more breeds.
• Calories intake varies from dog to dog, even within one single breed: This variation can be due to the habits of dogs and its owners (whether sedentary or active) or the age of the dog, and this will affect the amount of calories the dog requires too.