Title:

Poop Tracker: Predictive Modeling of Dog Movements Through Sensors

Dogs poop a lot, and leaving poop on the ground poses environmental and public health concerns and may result in fines in certain municipalities. Our project aims to create a system consisting of an app connected via Bluetooth to a collar equipped with an Arduino microcontroller. The Arduino uses a machine learning model to predict and alert the owner of defecation in real time, even when the owner is not physically present. Training data was acquired with a test app connected to a centralized cloud-based database. The data was used to train a machine learning model, which was then deployed onto the microcontroller for on-device inference. Whenever defecation is detected, the app displays the location on a map, allowing the owner to pick up and dispose of the poop.

Name

Comment*

Comments are viewable only by submitter

Captcha*