Kevin Shao

DawgSense

Your Dog's Emotions, Understood and Managed

HCI AR VR

Collaborators: Fei Huang, Adam Ji, Lirong Que, Negar Torkmanzadeh

Introduction

Many dog owners struggle to effectively respond to and manage their dogs' emotional and reactive behaviors, particularly in high-stress or unexpected situations such as barking at other dog or behaving abnormally in unfamiliar environments. This issue stems from a lack of tools to help owners interpret and address their dogs’ behaviors in real-time, compounded by limited opportunities to learn from these situations to prevent recurrence. This creates challenges in both immediate intervention and long-term behavior management, leaving owners feeling unprepared and dogs more anxious.

How Does It Work?

Our approach emphasizes equipping dog owners with the means to respond to immediate behavioral challenges with Augmented reality (AR) and reflect on those scenarios for deeper understanding with Virtual reality (VR). By focusing on both the reactive and proactive aspects of behavior management, the design ensures that owners are not only able to address a situation as it arises but also gain insights into potential triggers and alternatives for future prevention. This dual focus helps build owners’ confidence in navigating complex behavioral challenges while fostering a stronger connection with their pets.

Key Functionalities

Our two key functionalities—supporting in-the-moment crisis response and practicing skills to interpret emotional cues—align perfectly with DawgSense's dual AR and VR capabilities. The AR mode on Orion glasses enables real-time detection of behavioral cues and crisis triggers during live interactions with the dog, guiding the user to de-escalate or manage situations effectively as they happen. The VR practice mode enables the reflective learning by allowing users to replay past crises and revisit previously encountered scenarios, learning to identify and interpret signs of anxiety, stress, or aggression in a safe environment. The synergy between our focused tasks and design stems from how AR and VR complement each other, creating a well-rounded system that not only enhances immediate responses but also builds dog owners' foundational knowledge and confidence in understanding and managing dog behavior in various real-world contexts.

#1 Support In-The-Moment Response to the Crisis

**Monitoring and Real-Time Analysis Initialization:** Upon activating the monitoring function (AR mode), a pop-up notification “Analyzing in Real-Time” and a visual clue of the dog’s tail movement status appear. This assures the user that the system is actively scanning their dog’s behaviors and environmental triggers.

**Crisis Detection and Warning Message:** When unusual behavior is detected, such as rapid tail wagging or stiff posture, the system can recognize these early warning signs and alert the user with a warning message “Unusual Behavior Detected.” This message is accompanied by a prompt asking the user to confirm the system’s detection.

**Suggested Intervention and Crisis De-escalation:** Once a crisis is confirmed, the system presents a series of tailored suggestions to help de-escalate the situation.

**Post-Intervention Feedback and Confirmation:** Once the suggested intervention is completed, the system prompts the user to evaluate their dog’s current state to adapt dynamically to the outcome, offering further suggestions if necessary.

#2 Practice Skills Interpreting Dog’s Emotional Triggers and Responses to Prevent Crisis from Occurring

**Scenario Selection and Initial Setup:** Upon launching the system to practice mode (VR mode), users can choose a crisis scenario based on previously recorded incidents.

**Emotion and Trigger Recognition in Simulated Videos:** Once a scenario is selected, users watch a simulated VR video of the incident, observing the dog’s behavior in real-time. The system then prompts users to identify the dog’s emotional state, such as nervousness or fear, and the potential triggers causing these emotions.

**Feedback and Learning Support:** After users submit their responses, the system provides immediate feedback, highlighting whether the correct emotion and trigger were identified. If any triggers were missed, the system offers a list of potential triggers with explanations to enhance user understanding and learning.

**Performance Report and Improvement Suggestions:** At the end of the session, users receive a personalized report detailing their performance. The report highlights key achievements and offers suggestions for improvement.

Design Research and Key Insights

The goal of our design research was to understand the challenges dog owners face in interpreting and managing their pets' emotional and reactive behaviors. We aimed to identify gaps in current solutions and explore how real-time emotional monitoring could enhance behavior management. Our method was to conduct semi-structured interviews with three participants: a dog owner, a veterinarian, and a dog trainer, each providing unique perspectives on dog behavior, training, and management. These participants were selected for their diverse experiences, ensuring a holistic view of the challenges faced. Semi-structured interviews were chosen to allow flexibility in gathering in-depth insights into dog's emotional cues, behavioral management, and training needs. This approach provided actionable insights that informed our design decisions.

Real-Time Emotional Monitoring Enables Timely Interventions

“It’s hard to know what’s going on in her mind until it’s too late, and by then, she’s already reacting.”

The quote from the dog trainer revealed that dog owners struggle to recognize early signs of stress or anxiety in their pets, often leading to delayed responses that exacerbate reactive behaviors. This highlighted the importance of real-time tools for detecting subtle cues. Incorporating this into our design, we prioritized features like AR-driven real-time behavior analysis to guide owners in preventing crises before they escalate.

Owner Education Requires Long-Term, Reflective Approaches

“Dog owners need to learn how to manage their own emotions since stress can transfer to pets.”

The quote from the veterinarian highlighted the need for comprehensive education on dog behavior. This insight shaped our focus on combining immediate support with long-term learning tools, such as VR scenarios for reflective practice, helping owners understand their dog's triggers and manage interactions proactively.

Iterative Design and Key Insights

Our iterative design process started with low-fidelity sketches to establish basic design concepts based on our design research. We evolved these sketches into storyboards and task walkthroughs to convey context and illustrate interactions, focusing on two primary tasks: supporting in-the-moment crisis responses and practicing behavioral interpretation. This led to the development of paper prototypes and heuristic evaluations, where we identified crucial usability issues such as the absence of a back/undo button and confusing visual indicators. Usability testing with target users further revealed problems like insufficient feedback, complex task flows, and unclear navigation, which informed our refinement in the digital mockup stage.

Reflecting on this process, the iterative nature of our design approach—from prototyping to user testing—was instrumental in uncovering and addressing user needs effectively, making the final design both intuitive and effective. Each cycle provided critical insights that directly shaped subsequent revisions, culminating in a design that helps dog owners manage and understand their pets' behavior both in real time and through reflective learning.