MyWaterPal | Devpost

banner

Inspiration

For such a ubiquitous aspect of our lives, there exists such a massive disconnect between how much water we think we consume, and the reality of how much we actually consume. A poll from Water UK shows that UK adults believe they use less than 20 litres of water daily [1], when in reality that figure is closer to 150 litres a day in a single-person home [2]!

That's only considering household water usage too - i.e. bathrooms, appliances, etc. When you start to take into account the water utilised to:

Produce the very food you eat
Ship you the clothes you buy
Manufacture the products you purchase

It can be dizzying to try get a handle on the kind of impact you might be having on the world's water supply. It's the kind of impact that's really hard to put into context.

In an effort to help us start tracking this impact holistically, we built MyWaterPal. A software solution with hardware integration, giving people the tools to start tracking their water usage, in every aspect of their lives.

How Does MyWaterPal work?

First and foremost MyWaterPal allows you to accurately track your water usage from a variety of sources. These sources are grouped into two broad categories:

Recurring (through Sensors/Flow Meters) - this includes water used in the home at taps, in the shower etc.
One-Off (Purchases) - this includes water used in the production and transport of clothing and other products you have bought.

The app maintains a list of sources of water use, and under each source the app lists when water is used at this source and how much.

For purchases, this is achieved by taking a photo in the camera tab of the app. The Claude AI assistant is then used to assess what the captured item is and estimate how much water was used in its production and transport.

For other sources the app provides an API and, for each source, a unique id which can be used by external sensors to upload water usage data to the app. In our submission video, we demonstrate the operation of a water flow meter which sends data in this way via an ESP32 microcontroller board, which cost only ~£10 or ~$13 to build. This acts as a proof-of-concept for affordable "smart" water sensors across the home, which seamlessly interface with the MyWaterPal app.

How we built it

App : The MyWaterPal app was built using React Native and Expo. The backend is comprised of two AWS services: Lambda (API endpoints) and DynamoDB (NoSQL database).

feature-slide

(You can download the .apk file at the bottom of this submission, under "Try It Out")

Sensor Demo : The sensor was built using a 1/2" flow meter, an ESP32 microcontroller board, some 1/2" copper tubing and fittings. Code to host the web server, and process the incoming data from the flow meter, was written in C++ using the Arduino IDE.

Challenges we ran into

We have some issues building the .apk file for the app with eas build, in particular the time it took to build (we used Free Tier) and the app crashing when built, even while running fine on the development server. We eventually resolved these issues and, in the process, learned to better debug expo projects.

Accomplishments that we're proud of

We're proud of building an end-to-end system which integrates hardware and software with a good user experience and clear application. We're proud of building an effective demo with an ESP32 microcontroller, which we'd never used in a project before. Lastly, we're proud of having learned about and worked towards a water-sustainable future.

What we learned

By building the app, we learned how to capture and process images using expo-camera. We learned how to improve the user experience with splash screens, landing pages, easy refresh and skeleton loaders in react-native, which we did not have in previous projects. Lastly, we learned how to generate QR codes on the client side, and how to use the Anthropic API to interact with Claude.

In building the flow sensor demo, we learned how to program and host a web server on an ESP32 microcontroller, as well as how to process data from a flow meter (calculating flow rate from electrical pulses).

What's next for MyWaterPal

A mechanism for sharing videos and short articles on how to reduce your water usage from particular sources.
AI/algorithmically-generated and/or user-generated tips based on water usage data
Custom user-created water usage sources
Alternative power source for the flow sensor demo, so in practice it could be used without a USB connection.