Ideaz

Here’s the uncomfortable truth: most mid-size buildings don’t have a “water supply” problem—they have a visibility problem. Litres vanish via silent toilet leaks, rooftop tanks spill for minutes, and pumps run long—often during diesel-generator hours. By the time anyone notices, the tanker bill (and tempers) is up.

Why this matters to me

We’ve all paid inflated bills because of leaks or overflows no one caught in time. I’ve seen watchmen “measure” tanks with sticks and late-night guesswork on WhatsApp. I want a system that catches waste before it becomes a bill.

Who it benefits

• Residents/Tenants: Lower, steadier water and DG costs; fewer emergency tankers.

• Facility Teams/Security: Clear, actionable alerts instead of guesswork.

• Owners/Associations: Proof of savings, fewer complaints, longer asset life.

• Environment: Fewer tanker trips and diesel hours; less groundwater stress.

Current solutions & why they aren’t feasible

• Manual rounds & paper logs: Infrequent, error-prone, provide no early warning or audit trail.

• Float switches alone: Fail silently; can’t quantify losses or alert staff.

• Full BMS/SCADA installs: High capex, vendor lock-in, long deployments—overkill for 1–5 block sites.

• Smart meters everywhere: Accurate but costly to blanket; plumbing retrofits disrupt service and slow payback.

What AquaPulse does

A retrofit IoT + AI layer that makes water systems observable and actionable—without a heavy BMS.

• Level Watch: Waterproof ultrasonic sensor learns fill/empty rhythm; warns before overflow (“3 min to spill—cut pump?”).

• Pump Sentinel: Energy-metering smart plug reads power patterns to catch stuck floats/overrides; optional safe auto-cutoff.

• Leak Scout: Main-line flow sensor learns the night baseline; sustained trickle ⇒ leak alert with estimated loss.

Technical details (MVP)

• Hardware: ESP32 MCU; JSN-SR04T (ultrasonic), YF-S201 (flow), Wi-Fi energy-metering smart plug; 5V/12V PSU; IP65 enclosure.

• Firmware/Comms: ESP-IDF/Arduino; MQTT over TLS; per-device keys; last-will for sensor-health pings.

• Cloud/App: Rules engine (Node/Go), time-series DB (Timescale/Influx), web dashboard + Telegram/WhatsApp bot.

• Analytics: Rolling median + change-point detection; pump state (HMM-style); hysteresis to prevent alert spam; maintenance mode.

• Security: TLS, signed configs, least-retention of raw data.