Bird Boxers

• The existing decline in bird populations can be mainly traced back to several human factors. For example, tree swallow populations face severe declines largely attributable to rampant deforestation. The removal of forests destroys their natural habitats, disrupting the ecological balance, and affecting the availability of nesting sites and the insects that birds feed on. This points to a bigger issue of environmental damage as birds are natural pest predators that help maintain a balance in insect populations. When bird populations decline, the variety of insects consumed, flowers pollinated, and seeds dispersed may also decrease. Protecting bird populations is crucial not only for maintaining biodiversity but also for preserving the natural processes that sustain our environment. 

• Swallow boxes already exist in local Fremont parks. However, they lack the innovative technology that our Birdboxers team works to produce. In order to track useful environmental data regarding bird population dynamics, or even behavior and nesting details, Fremont parks recruit volunteers to open up each swallow box one by one. Not only is this process outdated and time-consuming, but it is disturbing to the birds themselves. Interfering with these habitats can upset birds especially in periods of egg incubation and chick development, introducing what seems like a threat from a predator. So, our technology, which will store data remotely in the form of photos and apply machine learning algorithms to learn about these fascinating species, helps to cultivate an understanding of the impacts of human activity on birds in a way that does not harm them. Given the threat of deforestation and climate change on tree swallow breeding and nesting habits, as well as the looming issues of climate change, pollution, and urbanization on existing migration cycles, the data we provide is incredibly valuable.

•  Our project has had the opportunity to attend some of the most significant events in our city, such as the Engineering Expo, which showcased more than 300 student engineering projects from various schools and attracted over 900 attendees to the Fremont Downtown Event Center. We also participated in Fremont's Earth Day, an even bigger annual event hosted by our city. These events have not only raised environmental awareness among citizens in our community but also helped us connect with people who can assist us in enhancing our city's pest management and population conservation. For instance, we have gained the support of notable city officials in sustainability, including Elisabeth Huffmaster, a clean Water Educator at the Nature Learning Center under the City of Fremont: Department of Environmental Services, Nathan Ivy, a Service-Learning Waste Reduction Project Coordinator at Alameda County Office of Education, and Mayor Lily Mei, City of Fremont's mayor. This support has set our project on a hopeful trajectory for the future.

• The Birdboxers project creates Population-Tracking Bird boxes across the Bay Area using Raspberry Pi computers and infrared cameras (accompanied by infrared LEDs) to take photos marked with their timestamp inside the bird box every 3 minutes, using bash shell scripts with packages such as lib-camera stills. Further, we have configured our project to store these photos into a pcloud photos album, using rclone copying software! The Pi is able to do all of this through manually enabled connection to wifi and cron jobs, which allow the software to start every time the Pi boots up. As a protective measure, we have also installed heatsinks, 3D printed a custom case with ventilation, and configured the Pi to send us email notifications and subsequently shut down if it starts overheating, if the incoming voltage it receives (via solar power) is too low, or if it disconnects from WiFi. To ensure longevity, our pi shuts down weekly for some scheduled downtime: during this period it runs updates, clears its disk space, and emails us a report of its current workings! 

• To view the software involved with this project, visit our github: https://github.com/birdboxers/Birdbox-Software

  • Look at our "How We Innovate" section for more details!

• We currently use wood boxes built for Fremont's Lake Elizabeth by local boy scout troops to house our tech!

To view the software involved with this project, visit our Github: https://github.com/birdboxers/Birdbox-Software

We started our prototype work in January, went through about 9-10 versions, and faced immense issues- the software was malfunctioning, the Pi was overheating, the soldering came off, and the solar panel and tracking camera were disconnected. Why? Because nobody had done what we were doing before. There is still so much work to be done, and below you'll find a concise documentation of our efforts!

Version 0

• Ms. Huffmaster gave us a Raspberry Pi Pico and told us to detect motion with it- we had NO idea what to do. 

Version 0.2 

• Upgraded to Raspberry Pi 02w

• introduced the Pi NOIR camera

• Pooja and Sophie tried downloading Raspberry Pi OS but it would require a monitor which we didn't have at the time

• Annika proposed using Motioneye and got it to work on a monitor- but we scrapped it

Version 0.4

• Mr. Huffmaster introduced us to NatureWatch

• Raspberry Pi 02w kept randomly shutting off/camera wouldn't connect

• Pooja and Sophie figured out how to adjust settings on it (hosts its wifi, successfully connects to host website, takes pictures periodically + real-time recording)

Version 0.6

• Sophie and Pooja set up the Raspberry Pi 02w with Infrared LED (using stripped jumper cables, Infrared LED, and added GPIO pins)

• Lashya and Harshitha soldered the model together

Version 0.7

• Upgraded to Raspberry Pi 4 model B (we fried the 02w) 

• Pooja, Lashya, Harshitha, and Sophie prepared the model (using a portable charger) for the Engineering Expo and Fremont Earth Day- developed a presentable display using an iPad, sheet protectors, and images to show our community our work and mission!

Version 0.75

• Naturewatch only stores on its host website but we wanted to store it on the cloud

• Sophie and Annika researched other programs and methods to both store footage in the cloud but have a connection to wifi (host its own?) - Motioneye came up again

Version 0.8

• Trying to convert to solar power, but charging very inconsistent

• Made a custom case using Fusion 360 for the Raspberry Pi

• Moving to motion eye again, trying to think of the real park applications 

  • Motion eye sucked; we weren't able to find the Pi's IP address to SSH into it.. and we didn't have a monitor 

    • Lashya, Annika, Sophie, and Pooja tried to set up Motioneye

Version 0.9

• Got a monitor

• Switched to just flashing Raspberry PI OS (Bookworm, 64-bit) onto the Pi

• Configured bash script for LibCamera-stills to capture photos 

• Setup storage on the USB drive 

• Solar power is still inconsistent- it only works with a super small charger

• At 50 degrees celcius when it booted, we were not sure how to deal with overheating

Version 0.95

• Tried to configure libcamera stills but libcamera for youtube streaming 

  • Libcamera was not creating executable files and the command was not runnable… then libcamera stopped working since 2 applications couldn't use the camera at the same time

• Overheating issues

Version 1

• Libcamera stills photos shell script every 3 minutes, used rcline to link the pi to our pcloud

  • Installed rclone and wrote the scripts for the pi to take photos and stores them there via the internet every 3 minutes

• Lashya/Harshitha/Pooja set up our website!

•  Solar panel → battery pack → pi

• Got heatsinks to help with overheating

• Stopped working once outside for a few hours...

• Wrote a script to make it email us and shut down if the voltage is too low or if it was overheating, it does this every 5 mins using cron jobs   

  • Didn't execute properly so every time it booted up it immediately shut down... so we had to manually edit SD card on computer 

Version 2 (To be implemented at Lake Elizabeth)

• Big 20W solar panel not the tiny little 5W   

  • Bought solar panel manager to regulate output, attached some 18560 batteries and dc cable to connect solar panel to manager, then connected this to a battery pack for a backup power source

• Added weekly maintenance, it emails us a disk usage report and updates itself every Tuesday at 2 am (although this is difficult as the power supply is still unstable at night)

• Added Wifi connection to the emergency shutdown script- it emails us and shuts down if it loses wifi connection  

• Added a backup measure in case the Pi is unable to copy photos to rclone, it stores them locally on its microSD card

• There was an overheating issue and the pi shut down, so we added aluminum tape and extra holes to the case + birdbox

• Planning to add tarp for weather/artificial shade at NLC to help with overheating

• Added a software constraint for overheating; if the pi is above 65 celcius, it emails us and stops taking photos- when it reaches below 60 it takes them again. If it ever gets above 70 celcius, it shuts down.

• Configured the pi to send images to Pcloud storage, not google photos like earlier (we kept hitting a google security quota)

Version 3 (To be Implemented at Oliviera):

• Sai got rid of extra back and side wood for oliveira 

• Software Team switched to the Pi 2w, edited the software to ONLY include the photos script to not overburden the hardware, and made the entire system much more energy efficient!!

• Alexis Redesigned 3d print design to be more compact 

• Rewired LEDS (not soldered)

• staple gun for cable securing

• Sai got wood planks to support the box from underneath

Version N/A (Backyard Experimental Implementation; BEI Version)

• Hardware Team Implemented a cbrp box (unmodified) in backyard

• Hardware Team used velcro -> then superglue -> then construction glue/paste to secure the pi case

• Sai Screwed birdbox to plank & used staple gun cables (solar panel and manager and pi)

• Alexis nailed a metal overhang to keep water (rain) out of the system

• Pooja velcroed the solar manager (case made of wood) to the top of the box

• Constantly Changing- this model is where we test out our most innovative, new ideas (like recording audio or weather monitoring!)

Version 4 (To be Implemented at Weibel):

• Took inspiration from CBRP boxes dimensions

• Added the side door and hinge system

• Rope system for tree securing

• Changed our pi (downsized to 2w)

• Added an audio system with the other 2w

• we deleted the overheating script and wifi check controls

*See our 'Workshops' tab for more information!

• Our team has already perfected the art of mass-producing autonomous avian-monitoring nestboxes using periodic infrared imaging. Our goal as Youth in an environmental nonprofit is to empower the next generation. 

• We host workshops where CHILDREN can TRULY make a difference, programming and constructing their OWN nestboxes and thus fostering a love for technology and sustainability. 

Groups that have had their Nest Boxes sucessfully implemented for Data Collection in the real world:

• TRAAS Team, 1/20 Workshop

• Sri Harini, Mika, Sophie, Alexander, and Jake, 1/20 Workshop

• When we were testing our prototype with an "empty" box, we discovered three tree swallow eggs nestled beneath the grass inside it. The box had previously been occupied by a bluebird when we brought it from the NLC. Despite thinking the box was empty when we took it home, we were surprised to find the eggs inside.

• After researching we found that one possibility could've been: Tree Swallow Abandonment Followed by Bluebird Nesting: It's possible that the tree swallows initially built their nest and laid their eggs but abandoned them for some reason. Later, bluebirds might have taken over the nest box and built their nest on top of the abandoned tree swallow nest.

• It is possible that human activity led to the abandonment of the swallow's nest. This could be due to intentional disturbances, like nest checking or habitat disruption, or unintentional disturbances such as noise or human presence, which can be perceived as threats by parent birds, leading them to abandon their nests. However, further data and observations are needed to substantiate this claim. 

• Nevertheless, this is impetus for our work. Striving to eradicate the human interference that caused this tragic issue and the death of potential tree swallow embryos, we can appreciate and understand our natural world without disrupting it's balance and harmony.