Inspiration 💡

I believe everyone of us knows that around 76% i.e more than ¾th of our mother Earth consists of water bodies & rest is Land of which we are a part of. That also means that there are around 8-10x times more wild marine animals which reside in those water bodies.

Times have changed, so have we. The increase in pollution is leading to an extensive increase in water volume which is resulting in huge chaos across the globe. But, unfortunately, it's very disappointing to see few people disagreeing with the same. This is not only harming us, but also causing intensive harm to the marine world. Besides, because of the same, thousands if not millions of species have become extinct and the death toll is still on the rise. The species which used to be common are now being counted under the list of endangered ones. The death of those living objects is amplifying the garbage in the sea/ocean and stacking the same with ours. Keeping all in limelight, it's disturbing the ecosystem as well as the entire biodiversity chain.

There's a famous quote that says — "We are unanimously entangled with Mother Nature & her Children." We catch fish for food, but also there are a few anonymous group of people who hunt/poach them for their purpose irrespective of the fact that it's disturbing the entire ecological lifecycle as well as the food chain of those animals. This is not good, but there's nothing that can help us evade the same in a correct manner.

Coming back to the main idea, have you ever tried fishing? If yes, you must know how hard and time tedious it is. Or while at a beach holiday you unknowingly interacted with something you weren't supposed to, which resulted in unexpected & unwanted circumstances just like this - Box jellyfish: Australian teenager fatally stung on Queensland beach while messing with a jellyfish.

The unwanted circumstances are not just for you, it's for the fishes too. Most of the fishes don't want to be disturbed by human interventions, the smallest mistake by you could eventually take their lives away. Recent climate change has also impacted Marine life drastically. Keeping all this in mind we brought an innovative and unique solution — Aquastreet


What it does 🤔

Aquastreet has got both software & hardware-oriented approaches. Leveraging cutting-edge Machine Learning inside our web application, we help to solve the above problems. It is the best buddy for a fisherman, people interacting with the ocean, to the fishes, & to the earth. Our app sought to protect both Humans and endangered marine animals from each other. Our project consists of two main parts

  1. Hardware — We have a pocket-sized hardware device that is compromised of an Arduino Nano 33 BLE Sense microcontroller. The microcontroller has got more than 8 different sensors embedded in it & the most beautiful part is that it's dimension is even smaller than my Pinky finger. The microcontroller is responsible for collecting marine sounds near you through the embedded microphone therefore the user has to place it at a convenient location preferably under the boat (it works underwater) or very close to the sea level in order to function properly.


The hardware can be preferably placed under the boat/ship in a closed enclosure. After analyzing the sound from those marine animals, we generate a MEL frequency spectrum & after waveform analysis, it is classified & the animal is predicted.


  1. Web Application — The front-end is crafted with Reactjs & uses Bootstrap as the CSS. To ensure that it runs on mobile we have taken a mobile-first approach. Users can log on to our web application on a laptop or android (supports mobile view) where Firebase is leveraged for both Simple Sign-in/Register or SSO. Users are required to sync the hardware to the application using Bluetooth. Then users can start a journey where the app will detect and track nearby marine creatures and they will be shown on your device. Names of the fishes detected, their density, and whether they are harmful or not would be displayed on your device. It is totally real-time and users can see whether they are moving towards the fish or away from it. If a dangerous creature has been detected like a shark, The app immediately notifies the user before it's too late. Also, since it's happening in real-time, we also keep a track of the position of the vessel which is done by basic arithmetic operations with respect to the distance covered (Speed = Distance / Time). That position is converted into Geo-JSON & then send to Firebase where we store the same in collections. We used Mapbox API to make the interactive map and fetch the user's location and mapping nearby marine animals. Users can share the data manually with anyone. If endangered species are detected the app asks the user to leave without hurting the fish, where we primarily target to protect both human & marine biodiversity. All of it is shown in an interactive geolocation map on the front end.

The goal of Aquastreet is threefold :

  1. Increase fishing success rates and reduce fishing time.
  2. Protect humans from potentially dangerous marine creatures like sharks and prevent an unexpected attack.
  3. Protect endangered and other exotic fishes from humans. Even a slight mistake can kill them.


System Architecture ⚙️

System Architecture - architecture


Challenges we ran into 😤

Building this was a very rough and tedious process. We faced countless challenges. I'll list the main ones

  1. Mapbox has very little documentation. So it was initially very had to integrate it on the react frontend.
  2. Integrating APIs when we were integrating the modules into one.
  3. CORS !!!
  4. Collaborating in a virtual setting but we somehow managed to finish the project on time :)

Accomplishments that we're proud of ✨

We are really proud of our app and ourselves. Some accomplishments boosted our confidence to another level, and they were —

  1. Completing such a complex app within a 36-hour timeframe.
  2. Making our team's first-ever Ambiware (Software - Hardware) hack.
  3. Getting 89% accuracy on our marine identification machine learning model.
  4. Using Lambda Functions & AWS Rekognition in a perfect manner!

What we learned 🙌

We learned a lot during the event. It was a very fun experience. Some of the major things we grasped:

  1. Using Amazon services like AWS Rekognition and Lambda Functions.
  2. Working with Arduino and combining it with TensorFlow lite.
  3. Creating websites that run better on mobile (mobile view)
  4. Using Mapbox API for plotting data & geofencing.
  5. Team coordination skills (a lot 😂)

What's next for Aquastreet 🚀

We have much planned for Aquastreet for the future :—

  1. A proper mobile app made in either React Native or Flutter.
  2. Adding support for Amazon DynamoDB & SNS.
  3. Enhance scalability of the app and more accurate detection with more fish/marine-animal information.

Conclusion — We had a lot of fun while building this app and had a great time at the hackathon. We learned a lot too! Thanks for hosting such a nice event 💖


Share this project: