Paw Pulse

Our project seeks to better understand our pets emotional state while they are home alone in an effort to align their living environment in ways that support their emotional wellbeing.

Comment

Inspiration

We started with a desire to deepen our understanding and connection with pets. Guided by the principle that “technology should serve as a bridge between living beings,” we created Paw Pulse to explore how a pet’s emotions shift throughout the day and how various environments impact those emotional changes.

What it does

Paw Pulse utilizes Qualcomm’s AI technology to analyze a pet’s facial expressions, movements, and vital signals in real-time. By detecting the emotional states of pets—even when their owners are away—it allows owners to respond quickly to their pets’ needs and better support their well-being.

How we built it

  • We capture pets’ behaviors and physical indicators (e.g., posture, facial features) through camera feeds.
  • We apply Qualcomm’s AI models for real-time image and video stream analysis.
  • The analysis results are relayed to a mobile/web app, providing immediate emotional status updates and alerts for unusual activity.
  • We store and analyze data in a server-side database, enabling long-term tracking of each pet’s emotional patterns.

Challenges we ran into

  • Limited available datasets for accurately estimating pets’ emotional states.
  • The complexity of training models to account for differences in breed, temperament, and age.
  • Balancing hardware optimization and ensuring stable network performance for real-time analysis.

Accomplishments that we're proud of

  • Improved recognition accuracy by incorporating data from various species and breeds.
  • Enabled objective observation of pets’ emotional flow throughout the day, even without human interaction.
  • Designed a scalable system architecture that can be extended to shelters, zoos, and beyond.

What we learned

  • Integrating animal welfare with technology has vast potential, emphasizing how human and animal emotional environments influence one another.
  • It’s crucial to use AI in an ethically responsible way, beyond simply focusing on technical accuracy.
  • Analyzing non-human emotional states requires a fundamentally different approach compared to human recognition models.

What's next for Pawpulse

  • Collaborate with animal shelters to provide emotional profiles and improve adoptability matching.
  • Expand to monitor emotional landscapes in farms, zoos, and larger ecosystems.
  • Develop interactive features that automatically adjust environmental conditions—such as lighting, temperature, and noise—based on the emotional data collected.

Built With

Share this project:

Updates