Inspiration
Our team members understand the difficulties of taking care of our favorite furry friends as they get older. Watching as their vision and hearing goes away and it becomes more difficult for them to live their daily lives, we are always concerned about their safety and well-being. While we can always provide great care to them in our homes, leaving them alone is always a worry due to not being able to guide them in their journeys. That is why we developed PupSense, an embedded system focused on providing the best protection for our wise furry friends.
What it does
This hardware-oriented project focuses on utilizing Arduino and LiDaR sensors to track where the pet is within a home. We envision this project to be integrated in a harness that the dog can comfortably wear and utilizes different features to provide protection. Our Arduino system encorporates ultrasonic sensors and buzzers to alert the pet when an object walks by, and we utilize a Raspberry Pi with LiDaR to create a local mapping of the environment to help provide an alert as to when the dog it near some kind of obstacle. All of this information is integrated within an online dashboard you can access from home, containing a live camera feed and statistics of your pet such as a heartbeat monitor.
How we built it
For the physical components, we utilized an Arduino UNO R4 WIFI with ultrasonic and vibration sensors and buzzers to emit sound once a certain threshold is heard, with an Arduino sketch running to create audio alerts when an obstacle is in close range to the pet. For our Raspberry Pi + LiDAR setup, we used Ubuntu 20.04 with ROS2 architecture to run nodes to create local mappings of the environment using Cartographer mapping software, with our LiDAR being a Garmin LiDAR Lite v3. As for the web dashboard, we utilizing a Flask backend server for receiving webcam transmission from the Pi and processing sensor data. For the frontend, we used HTML and CSS to properly display all of the information, including a graph for the heartbeat rate, local mapping, and camera feed.
Challenges we ran into
Trying to properly utilize ROS2 architecture was especially difficult for us to us being first-time users of the software and configuring Ubuntu 20.04 on the Raspberry Pi in general. Its limited processing power and the older distribution made it much more difficult to implement ROS2 than anticipated, and we ran into many issues regarding performance that caused us to reevaluate what features to implement.
Accomplishments that we're proud of
We were able to get live video feed properly displaying on a Flask backend server, along with creating an Arduino alert system based on proximity sensors.
What we learned
We were able to learn about integrating simulations into ROS2 architecture and how to create local mapping systems. We also had some members learn about web development for the first time.
What's next for PupSense
We want to ensure the LiDAR is fully utilized and create proper local mapping systems using ROS2 architecture. We would also love to refine the dashboard to be able to show more information and statistics to the user. Lastly, we want to investigate more sensors to utilize within the Arduino system for more accurate feedback for guiding the dogs, such as vibration modules that help direct the dog away from obstacles.
Log in or sign up for Devpost to join the conversation.