Milestone 3

Introduction

After Milestone 2, our team has shifted focus to monitoring rather than separation anxiety, which we were initially focused on. In the previous milestone, we completed a diary study with 10 participants and 4 user interviews to better understand our problem space and user group. The diary studies and interviews both revealed that pet owners were more concerned with monitoring their pets in case of any accidents, emergencies, or health-related things rather than separation anxiety as we originally hypothesized. Based on our findings, we brainstormed different concepts that centered on monitoring and decided to move forward with our smart collar concept with some features of a camera as well. For this milestone, we took that concept, created low-fidelity prototypes, and wrote out different scenarios, and conducted user enactments.

Study Design

To start off our user enactments, we decided to pick one of our three concepts and ultimately ended up combining two of them - the smart collar and the camera. From there, we created a speed dating matrix. The speed dating matrix helped us plan out what scenarios we could test. The entire matrix we created is below with at least nine different scenarios.

After this, we created specific scenarios that we would like to test with users. We wrote scripts and sketches for 5 different scenarios that we then tested with 5 users from our target group of cat owners.

User Scenario Script 1:

Asking the owner to create a new profile for the cat and fill in basic info about their cat, give the users instructions on how to connect the collar and camera (optional)

Script/sketch:

User A has just purchased the Colcat kit. He receives the Colcat collar and downloads the Colcat App. He quickly goes through the onboarding process and learns the functionalities of the App and device. He connects the collar with the App and puts it on his cat. He plays around with the interfaces and tests if the tracking data looks accurate. He sees that the tracking data mainly includes two parts: emergency and health-related metrics. He can select exactly what he wants to track on and what emergency he wants to be notified of. He then selects sleeping/heart-rate/body temperature/exercise as the data he wants to track; and he selects ‘eating outside the meal time’ and ‘extreme heart rate’ as the emergency he wants to be notified of. He knows that the camera is an optional device he can connect to as well, so he pairs the home security camera he had with the app and uses it to monitor the cat.

User Scenario Script 2:

The collar alerts the owner of the changes in the cat’s behavior, and helps prevent emergencies from happening.

Script/sketch:

The owner is having chocolate ice cream and he receives a phone call from his boss that he has to attend an in person office meeting today. He’s in a hurry and forgets to put the ice cream back in the refrigerator. His cat notices the ice cream on the dining table and jumps to it. The cat smells the ice cream and decides to eat some. The collar detects the motion and notices the cat is eating something outside his dinner time. The collar sends the owner a notification that the cat is eating something and attaches a photo of the cat (which is taken by the home security camera). The owner views the image and reads the message from the app. He realizes the ice cream contains chocolate which is considered toxic for the cat, he calls his neighbor for help, in addition, he makes the collar ring, which annoys the cat and prevents a tragedy. He sees that the neighbor enters the apartment and gets rid of the ice cream, via the camera. The collar captures a drastic heart rate increase of the cat, which is caused by the neighbor.

User Scenario Script 3:

The cat gets anxious because there are loud construction noises outside the house. The owner gets a notification that the cat’s heart rate is high

Script/sketch:

The owner goes out for work and leaves the cat at home alone. He gets a notification that says the heart rate of his cat is abnormally high. He is worried about his cat and curious about what triggers this. He turns on the camera view from his App and watches the footage of his cat and wants to find out what happened. He hears from the camera that there are loud construction noises outside his apartment, and sees his cat was scared and hides below the couch. He feels a little bit relieved after knowing more about the situation.

User Scenario Script 4:

The cat is wearing the collar and the collar has been monitoring its health statistics the entire time. The cat owner will receive an alert as needed or determined by the collar.

Script/sketch:

User Scenario Script 5:

The collar’s battery is running low

Script/sketch:

The cat has been wearing the collar for almost a week, the battery life of the collar is low, so the “low-battery” light starts blinking in red. The collar sends a notification to the owner, and he reads it on his phone. While the owner gets home, he takes off the replaceable battery from the collar and puts the other battery (from the ColCat kit) on the collar. The red light is gone, and the smartphone app reminds the owner the battery is successfully replaced and the battery life is 100% and can last for another week. The owner put the empty battery in the charge station.

For each of these five scenarios, we had props set up. We created a prototype collar and wireframed screens to use in our user enactments. From there, we tested out each scenario with our 5 users. We completed two of the enactments with the class with paper screens and paper prototypes for a collar and camera. After we completed these two we took the feedback we had from these two enactments and implemented them in a low-fidelity wireframe on Figma. We created our wireframes in Figma to make it easier to test with. Next, we completed three more user enactments which we used a soft toy as our cat and used Figma mirror to use the wireframes. All of these yielded really great results detailed below.

Study Results

Scenario 1: Setting up the application and collar

Participants were generally able to connect with wifi and connect with the collar intuitively. Some users were not sure if they needed to connect the collar to wifi In addition to connecting the phone, so we need to re-evaluate the user flow for connection and determine whether wifi is the necessary type of connection. When connecting the camera, one participant wished that there was a screen to see the camera view so they can set the desired frame.

One of the steps in the set-up process Is to create your cat's profile and add them to your device. One participant believed this feature was for adding a picture of the cat rather than creating Its profile. The cat profile also Included factors such as characteristics describing the cat's personality. Users reported that this information was not useful, but that age, sex, and breed were Important to Include In the profile.

The set-up process also Included video Instructions for Installing the camera and collar. Users reported that they do not normally read Instructions, and instead of having videos that demonstrate each step of the process, we should consolidate the Instructions onto one page.

Users also selected the criteria they wanted to track about their cat and used the dashboard to view metrics. Multiple participants said that showing a heartrate visualization was confusing because typical users do not understand the visualization. One participant said, "I'm not a doctor, so I would not understand this". Many participants said to Include a "normal" benchmark for each metric so they understand when their cat Is below or above average for a category.

Scenario 2: The collar alerts the owner of the changes in the cat’s behavior, and helps prevent emergencies from happening.

When the user receives a notification that their cat is eating something toxic, many participants believed that the option to call your neighbor was too extreme, and that there were more ways to nudge your cat. The app also gave users a notification when their neighbor came to their house to take care of the situation, but participants were definitely alarmed that the notification said "someone Is In your house" and suggested a more general message like, "motion detected".

Some users also reported asked, "how does the device know that the cat Is eating something toxic?". They suggested that a more feasible notification would be that your cat Is eating In a different location than usual, and that would be an Indicator that your cat Is not eating something normal.

The app also needs to be more clear that the call and text option links to the phone and messaging apps on the user's phone rather than calling and texting via the app.

When the user opens the metrics dashboard to check the cat's health, many users reported not to use red except In case of emergencies.

Scenario 3: The cat gets anxious because there are loud construction noises outside the house. The owner gets a notification that the cat’s heart rate is high

Many users reported that the dashboard has a lot more Information than necessary upon first glance. They suggested having a summarized view and clicking on a category to expand to more detail. Users also liked the variability In the types of visualizations (e.g. bar graphs) on the dashboard.

Scenario 4: The collar’s battery is running low

Many users liked the ability to switch out the battery for the collar. There was conflicting feedback about how the batter protruded out from the collar. Some thought the design was clunky, while others thought It was more noticeable easier to take on and off. Some users were not sure If they needed to charge the whole collar or just the battery. Many participants said they wanted to know the battery level of both the current battery and the replacement battery to make sure the replacement would be able to work.

Participants had mixed feelings about the low battery notification. They wanted to see an option to Ignore the notification. The paper prototype was designed so that the user would need to open the notification and app in order to replace the battery, but one user reported, "I don’t think I will open the notification because it says the battery is low, I would just replace the battery without opening the app. The notification has already told me enough information". Our team needs to design the battery replacement process to be Independent of opening the app, In the case that the user just wants to replace the battery upon reading the notification.

Key Takeaways from Results

Project Scope:

The scope of our overall project remains the same, participants found the product to be useful to their needs. The most significant change in scope for the product was related to involving outside stakeholders like neighbors. The cat-owner relationship Is highly Individual, and cat owners did not see the need to call a neighbor unless there was an emergency. Most changes In project direction are focused on feature-level feedback.

User needs, design constraints, and design opportunities:

Provide a "normal" benchmark for each data visualization - users reported that having a visual Indicator of whether their cat Is above or below normal would be the easiest way to tell If the cat Is doing well or has any Issues
Avoid using extreme language or colors - using red In a data visualization often Indicates something alarming, and users did not react well. We also need to soften the language around notifying users of Issues. Users want to be able to speak directly to their cat through a microphone rather than call a neighbor, and do not want to see the message "someone Is In your house" If their neighbor does come to their house.
Reduce the number of steps In onboarding - because many participants reported that they do not read Instructions, our team should design the app so that all Instructions are available on one page and are able to be skipped. We should also allow the user to refer back to the Instructions at any point If they get stuck.
Provide more Information about battery life - users want to know the battery life of both the current and replacement battery so that they can avoid the situation of replacing a dead battery with another dead battery.

Ideation & Selection

Our ideation and selection have been conducted throughout the entire user enactment process. We mainly focused on and 'mobile application feature' variations. We ideated our mobile features and user flows on paper prototype at first. In order to select the 'final' features that can better fit target users' needs, we transitioned our paper prototype to digital low-fi prototype to create a more realistic experience for user enactment participants.

For better quality full image: https://miro.com/app/board/o9J_lu2BV5w=/?moveToWidget=3074457367607877374&cot=14

After analyzing findings from our user enactments, we further refined our original concept made some changes and removals based on the insights gained from the participants. We ideated multiple choices for our mobile app screens. In order to select our ideas more logically, we created the following criteria to evaluate our design concepts.

Utility: Does the feature help fulfill a specific need of target users?
Priority: Does the feature need to exist in order to function/address our primary needs?
Value: Do users recognize the value of utilizing this concept?
Uniqueness: How unique is the feature?
Demonstrability: Can the feature be easily demoed?
Support multiple cats: Does the product support the users who have multiple cats?

Here are some of the selections and updates we made, we classified them into 'add'; 'removal' and 'change':

ADD

Profile set-up for multiple cats: As we were testing our Scenario 3, we understood from our users that many cat owners have multiple cats and if they purchase multiple collars for their cats, they want to be able to connect each collar with the App individually to monitor all cats.

REMOVAL

Removal of Battery replace instruction: Durning our testing on 'The collar’s battery is running low' scenario, we understood that the replacing battery instruction is not very helpful and necessary for returning users, so we decided to remove the instruction screen.

CHANGE

Change of the dashboard layout: After testing Scenario 2, we heard feedback from our users that they want to clearly see the normal vs. abnormal metrics at one screen. We then updated the screen layout having two sections for abnormal & normal metrics on one screen.

5. System Proposal

The IoT solution we are proposing is ColCat, a smart wifi-enabled cat collar with sound, geolocation, motion, heart rate, and body temperature sensors. It effectively tracks the health metrics about cats and alleviates cat owners' worries while they have to leave the cat home alone.

Cat owners can view up-to-the-minute health data of their cats from the easy-to-use ColCat mobile application, receive alerts while their cats behave abnormally, and connect compatible home cameras and other smart home devices to the ColCat ecosystem to further understand the physical and physiological health of their cats.

5.1 Storyboard & Low fidelity mockups

Script 1. User onboarding process: how to pair ColCat with your smartphone and how to connect additional smart home devices (such as home camera and smart cat litter box) to help track your cat.

Script 2. Cat home alone, attempting to eat ice cream with chocolate, which is toxic. This dangerous behavior is detected by the ColCat collar, cat owner gets an alert that the cat is eating something at a different spot and outside of the cat's regular mealtime. Cat owner views the photo captured by the home camera which is connected to ColCat and confirmed the cat is in danger. Then the cat owner controls the collar via the app and makes it ring, the noise stops the cat from trying to eat the ice cream. He then sends an emergency message to the emergency contacts he previously added, and his friend, who lives nearby, goes to the cat owner's apartment and saves the cat.

Script 3. The Cat owner is outside, he receives an alert that the cat's heart rate is higher than normal, so opens the app and checks the metric, the heart rate is high, so he opens the camera view. Based on the ColCat analysis, his cat is scared by some loud noises outside the apartment, so the cat is hiding beneath the couch. He feels relieved.

For better quality full image: https://miro.com/app/board/o9J_lu2BV5w=/?moveToWidget=3074457367686849503&cot=14

5.2 System Components & Features

The ColCat IoT system is built upon 4 major components, the ColCat Collar, ColCat Mobile App, additional compatible smart home devices, and the Vet (cat hospital) Network.

In our vision, the ColCat Collar will be implemented with low power consumption sensors including sound, geolocation, motions, heart rate, and body temperature. And with smart algorithms, all sensors with work together to analyze how the cat is feeling and doing, automatically. For instance, the sound unit will record sounds produced during the cat's routine activities like meowing/eating/drinking/using litter boxes, the motion and geolocation sensors will monitor what the cat is doing, and help home cameras to trace where the cat is, etc...
All data collected by ColCat collar will be sent to the Cloud for processing and cat owner will see the breakdown health metrics information directly from their phone. Cat Owners will receive alerts about abnormal situations of their cats, weekly reports of their cats' physical and mental health.
Cat owners can connect compatible smart home and cat IoT products to ColCat collar through the ColCat app, which will produce a more accurate and useful cat tracking system for the owner. For example, cat owners can connect the smart litter box or cat feeder, so the ColCat collar can better track the cat's eating/urinating behaviors. Cat owners can connect Alexa home cameras and speakers to ColCat, so they can better monitor the cat or talk to the cat even if they are not home.
[Future Feature] If the cat is sick and needs to see a vet, the cat owner can share some data via the built-in health informatics exchange (HIE) system with the pet hospitals, so the vet can better understand what happens to the cat.

5.3 Technical Limits

From our previous research, cat owners find it hard to know how their cat feels when they are not home. In our IoT solution, we would really want the collar to track the cat's emotion, which seems to be a technical challenge. Our proposal is to use the built-in sound sensor to record the cat's meowing activities and use the home camera to capture photos of the cat's facial expressions, then dump the data into ColCat sophisticated machine learning and artificial intelligence [not existing yet], so the system can infer the emotional changes of the cat.

6. Demo Proposal

Based on the time constraint and the skill sets of the team members, we aim to design a mid to high fidelity prototype of the ColCat collar. We will purchase a standard nylon cat collar and add different sensors to the collar to demonstrate how the ColCat collar product looks like. Specifically, we will use materials like play-doh and paper to create the sound, heart rate, geolocation, motion, and body temperature sensors, and 2 rechargeable batteries. We will also use a mini-led bulb to design the wifi connectivity indicator.

For our mobile app, we will design a high-fidelity prototype. We will create an interactive mockup in Figma. The main screens we are going to design are onboarding pages, alerts inbox, dashboard, health metric detail page, and a few device control pages.

To demonstrate ColCat's ability to work with additional smart home devices, we will use a home camera and a smart speaker to present a scene in which the cat owner wants to see their cat and chat with the cat while they are outside. This demo will rely on some video editing techniques in our final video.

Conclusion

Our solution successfully meets the needs of cat owners who want to monitor their pets and our scope still remains relevant to our target users. The key takeaways from our research phase Involve changes to product features and design. Our team needs to evaluate the fundamental user flow of onboarding the user to the product as well as the level of detail and type of Information present In the metrics dashboard. The team also needs to revisit some technical limits of the application regarding the type of sensors and Information that IoT will be able to track and the type of network connections necessary to Implement the product. Our focus going forward will be to design a high-fidelity prototype that will Incorporate this design feedback to create a more Individual and customizable approach for cat owners to monitor their cats.

Page updated

Report abuse