Cookie - "The Boss"
we had to build our own board
curious chickens take debugging to a whole new level
What are those chickens up to?
It all started when my wife and I got chickens a few years ago. As any new chicken owner will tell you, it's pretty exciting to think about when you will get your first egg. After all, you've been caring for these little animals for several months by this point - not to mention the time spent building the coop! For our flock, the nesting area was inside the coop where we could not see the birds. One day I was sitting in the back yard watching the chickens (this is called "chicken-tv") and I started wondering what were they doing in there?
What it does
After playing with a few approaches, we settled on detecting when the chicken is on the nest and for how long. We use this information to decide when the chicken has laid an egg. The nesting data is combined with other sensor data inside the coop as well as data from external APIs for weather conditions outside the coop. Together we have a wholistic picture of the conditions which are most favorable for poultry production and can use this data to help optimize the best laying conditions for each chicken. Eventually we'd like to be able to say which chicken performs best in which zip code and with what other breeds of birds.
We've collected over 250k data points so far and are just about to install our first device in California where we think we can help small farmers make their pasture raised poultry more efficient.
How we built it
Lots of late nights and iterations! There are two main sections of work A) the device and B) everything else. We are fortunate to have two people on this project - one with proficiency in each area!
The device collects the data and sends it to AWS IoT where we use their rules to store the messages in DynamoDB and forward them to Lambda for further processing. Ultimately the data is collected in a relational database which is accessible from the web for visualizations and for Alexa "Alexa, ask Chicken Scoop how many eggs today?".
Challenges I ran into
Hardware is hard - lots of nuance in every piece of our boards and hardware architecture. It turns out that "clean" power is super important for devices - who knew?!
Other challenges were the chickens! They were scratching at the wires for our prototype boards, laying eggs in weird places - nothing like bringing animals into contact with technology to keep an engineer humble!
Accomplishments that I'm proud of
Our main coop has 90% accuracy in detecting egg production - we're very proud of that! Plus we've been able to help people identify good and bad behavior in their chickens. It turns out that this data tells us more about chickens than just when they lay eggs. We've also been able to help a local school bring their chickens into the curriculum with a data-based approach. We are excited to keep learning!
What's next for Chicken Scoop
We'll be working with the University of California to monitor one of their coops and see what we can learn about egg production from their flock. We're currently looking into the commercial viability of this product - maybe we'll have a Chicken Scoop in your neighborhood some day!