Demo: https://aquashift-green-shift-hacks-2026.vercel.app
Source code: https://github.com/samertall/tools/tree/main/hackathons/aquashift-green-shift-2026
Inspiration
Sustainable decisions become easier when people can see what a practical change might mean for them. Household water-conservation advice often lacks personal context: a short shower and a repaired tap may have quite different potential effects depending on usage. AquaShift puts editable assumptions and transparent math in front of users, while separating projections from observed water meter data.
What it does
AquaShift is a working, responsive web tool for household water conservation. A user specifies household size, shower frequency, shower duration, fixture flows, and dripping tap rate. They can then toggle four actions: shorten showers, install a lower-flow showerhead, turn off the tap while brushing teeth, and fix a leaking tap. The page immediately calculates modeled liters/month before and after and breaks down which activity contributes to the estimate. Separately, users can log cumulative meter readings, calculate interval daily averages, record daily habits, export the data in CSV and clear stored records.
How it was built
The application uses HTML5, CSS3, native JavaScript and browser localStorage, with no backend, signup, analytics or third-party API. The pure scenario calculation engine is tested using Node's built-in test runner. A Playwright/Chromium smoke test verifies reactive interactions, meter computation, habit recording and mobile layout. The model assumes 30.44 days/month, two toothbrushing sessions per person/day, and 0.25 mL/drip. When two shower changes are selected, their effects are combined in one scenario, avoiding double counting.
Environmental benefit
The app is designed to reduce avoidable household water use by letting users explore the size and relative importance of plausible interventions. Under a clearly labeled illustrative two-person household configuration, selecting two-minute shorter showers, turning off the tap while brushing and repairing a dripping tap models approximately 2,447 liters/month of potential water savings across the modeled activities. This is a calculation, not a verified environmental impact result, and it does not represent the entire water bill.
Challenges
The main challenge was communicating impact without pretending that generic assumptions are a measured fact. AquaShift keeps estimates separate from the water-meter section and explains the limitations directly in the user interface. It also handles contradictory meter readings, missing readings, and local storage failures.
What we learned
Personalized scenarios can make conservation actions easier to prioritize, but transparent assumptions, accessible UX, and honest limitations are as important as the headline number.
What's next
Potential future work: optional household baseline calibration, richer interventions, local tariff assumptions, water meter integrations where permissioned and available, longitudinal studies, and a stronger evidence-based attribution model. None of these are claimed as completed.
Technologies
HTML, CSS, JavaScript, ES modules, localStorage, Node.js, Playwright, Chromium, Python/Pillow and FFmpeg (demo assets).
AI tools used
OpenAI GPT-6 assisted with product scoping, UI implementation, scenario calculation code, testing, documentation and demo materials. Results were checked using automated tests. No external LLM or API is part of the runtime app.
Testing instructions
Visit https://aquashift-green-shift-hacks-2026.vercel.app . Toggle an action, change household size and confirm the estimate updates. In the meter panel, add cumulative readings of 100.000 cubic meters for 2026-10-01 and 102.500 cubic meters for 2026-10-11. The measured interval should be 2,500 liters and 250 L/day. Mark a habit and reload; your local record should remain on that device. Try CSV export and the reset controls. The automated Node tests cover seven model and meter cases.
Limitations
Savings are illustrative scenario differences, not guaranteed reductions, bills, carbon-emission estimates or scientifically attributed measured outcomes. Only showers, toothbrushing and one dripping tap are modeled, so this is not whole-home consumption. Local storage is device- and browser-specific; the prototype does not sync across devices or connect to smart meters. No user field study was carried out.
Team and original work
Samer Tallauze — individual project author. Source is an original prototype developed for this event period, with AI assistance acknowledged. No external datasets or stock media were required.
Interactive demo walkthrough for judges
- Open the working prototype at https://aquashift-green-shift-hacks-2026.vercel.app and inspect the initial modeled monthly water-use baseline.
- Toggle shorter showers, efficient showerhead, tap-off brushing, and leak repair. Compare baseline, improved scenario, and savings breakdown. Change household size, shower frequency, and fixture flow to see how assumptions affect outcomes.
- Log two cumulative water meter readings: 100.000 m³ on 2026-10-01 and 102.500 m³ on 2026-10-11. The panel reports 2,500 liters used across 10 days, averaging 250 L/day. These are illustrative demo values, not claimed user measurements.
- Mark a daily conservation habit, reload to check local persistence, and export meter and habit records as CSV.
- Inspect the transparent source code and run the seven calculations/meter tests from the GitHub project. The source and README are at https://github.com/samertall/tools/tree/main/hackathons/aquashift-green-shift-2026 .
This step-by-step interactive walkthrough is provided to meet the event's demo-video-or-walkthrough requirement. A separately generated 30-second MP4 overview is available in the downloadable project package; no public YouTube/Vimeo URL is claimed.
Log in or sign up for Devpost to join the conversation.