AquaPulse — Urban Freshwater Bio-Sentinel & Citizen Telemetry Intelligence

💡 Inspiration

Urban freshwater ecosystems support over 55% of the world's population, supplying municipal drinking reserves, cooling city microclimates, and fostering rich riparian biodiversity. Yet over 70% of urban rivers and streams face chronic ecological degradation from sudden industrial runoff, untreated stormwater discharges, and warming thermal plumes.

Standard municipal water monitoring relies almost entirely on sporadic chemical grab-sampling. These sparse tests suffer from three fatal flaws:

1. Temporal Blindspots: Nocturnal industrial dumps or sudden stormwater pulses dissipate before field technicians arrive with sample vials.
2. Missing the "One Health" Triad: Chemical data is analyzed in clinical isolation, completely ignoring how water degradation directly correlates with zoonotic disease vectors (such as Culex mosquito breeding in stagnant hypoxic eddies) and domestic pet/human dermal toxicity.
3. Citizen Disenfranchisement: Stream quality reports are buried in dense regulatory PDF documents, preventing local riverside communities and student researchers from participating in stream stewardship.

We were inspired to build AquaPulse: a unified, bio-inspired urban freshwater telemetry and citizen intelligence platform that fuses real-time physicochemical sensors with benthic macroinvertebrate bio-indicators and participatory community science.

💧 What It Does

AquaPulse transforms complex limnological and ecological formulas into an intuitive, accessible, and actionable dashboard:

• Composite Ecological Health Index (EHI): A unified 0–100 score synthesizing Benson-Krause Dissolved Oxygen saturation percentages, pH deviations, turbidity penalties, and biological biotic indices.
• Hilsenhoff Family Biotic Index (FBI) Calculation Engine: Uses benthic macroinvertebrates (such as pollution-intolerant Mayfly and Stonefly nymphs versus hyper-tolerant Bloodworms and Tubifex worms) weighted by standardized tolerance values ($t_i \in [0, 10]$) to diagnose chronic organic pollution weeks before chemical thresholds collapse.
• Benson-Krause Temperature-Compensated Oxygen Solubility: Accurately computes theoretical oxygen saturation ($DO_{\text{sat}}$) based on ambient water temperature and barometric pressure.
• One Health Multi-Vector Risk Radar: Continuously evaluates risks across human dermal contact, companion animal ingestion toxicity, and vector-borne pathogen proliferation.
• Interactive Global Catchment Simulator: Live baselines for 5 iconic global waterways—Barton Creek (Austin), River Spree (Berlin), River Thames (London), Sumida River (Tokyo), and Bellandur Lake (Bengaluru)—with real-time parameter tweaking.
• Citizen Science Field Telemetry Console: Enables residents and student volunteers to log geotagged macroinvertebrate tallies, attach stream clarity observations, and receive cryptographic verification and stewardship badges.
• Municipal Remediation Planner: Evaluates nature-based solutions (Macrophyte Floating Wetlands, Riparian Buffer Strips, Micro-bubble Solar Aerators, Benthic Bio-char Beds) with projected EHI recovery curves, ROI metrics, and implementation timelines.
• Sensory Acoustics: Real-time water droplet and sonar chime synthesis powered natively by the Web Audio API without external audio files.

🛠️ How We Built It

• Scientific Modeling Core (Python 3.10+): Built a mathematical engine implementing the Hilsenhoff FBI formulation, the Benson-Krause polynomial for dissolved oxygen solubility in water, multi-metric EHI aggregation, and the One Health risk matrix.
• Automated Verification: Engineered an automated unit test suite (test_aquapulse.py) validating mathematical edge cases, USGS reference calibrations, and organic pollution thresholds with 100% pass rates.
• Tactile Marshmallow Glassmorphic UI: Designed a responsive interface using Plus Jakarta Sans and JetBrains Mono typography, custom bioluminescent gradients, smooth 28px marshmallow cards, and reactive visual gauges.
• Native Web Audio Engine: Implemented dynamic harmonic droplet synthesis using native Web Audio oscillators and exponential gain envelopes.
• Video & Telemetry Production: Scripted an automated 1080p Playwright recording walkthrough synchronized with neural voiceover narration via edge-tts and compiled using ffmpeg.
• Deployment & Hosting: Global edge deployment on Vercel with clean routing, and complete open-source version control with chronological git history on GitHub.

🧗 Challenges We Ran Into

• Limnological Temperature Calibration: Modeling dissolved oxygen requires accounting for the non-linear relationship between water temperature and gas solubility. Using standard linear approximations caused wild errors at higher summer temperatures. Implementing the full 4th-order polynomial Benson-Krause formula solved this with USGS-grade precision.
• Biotic Index Sensitivity: Benthic macroinvertebrates have varying population sizes. If only a single tolerant organism was reported in a tiny sample, naive averaging could severely skew the entire catchment health. We introduced weighted sample density thresholds to ensure statistically sound biotic scoring.

🏆 Accomplishments That We're Proud Of

• Six out of six passing unit tests executing in under two milliseconds.
• Seamless bridge between rigorous environmental science and delightful, accessible consumer UI.
• Interactive catchment simulation allowing judges to test extreme pollution scenarios and witness instantaneous One Health risk recalculations.
• Zero external audio assets—complete ambient sound generation rendered directly through math in the browser.

📚 What We Learned

• How bio-indicators act as continuous living sensors: unlike chemical grab samples that capture a single instant in time, macroinvertebrate communities integrate the cumulative impact of environmental stressors over months.
• The vital importance of the One Health framework in urban planning: degraded water bodies are not merely aesthetic issues—they are direct epidemiological drivers of vector-borne illnesses and municipal health crises.

🚀 What's Next For AquaPulse

• Computer Vision Macroinvertebrate Identification: Integrating a mobile camera ML model capable of classifying live benthic nymph specimens directly in the field.
• LoRaWAN Stream Sensor Nodes: Deploying low-power, solar-harvesting buoyant sensor buoys for continuous telemetry streaming directly into the AquaPulse catchment mesh.

Built With

• css3
• git
• github
• html5
• javascript
• playwright
• python
• unittest
• vercel
• webaudioapi