IoT Sensor Network

Overview

This project began as a practical prototype: collect a few environmental signals (temperature/humidity, basic telemetry) from a cheap microcontroller sensor node and visualise the data locally.

Only Iteration 1 was built. A more ambitious "Iteration 2" design was planned (long-range comms, power budgets, broader coverage) but never progressed beyond concept.

Iteration 1 (Built)

What it does

• A sensor node captures basic readings at a fixed interval
• Data is transmitted over Wi-Fi to a local collector (Raspberry Pi)
• Readings are stored locally and graphed in a dashboard

Why I built it

• Learn the full path from "hardware signal" → "data ingestion" → "storage" → "visualisation"
• Understand real constraints (Wi-Fi reliability, noise, calibration, time sync)

Practical takeaways

• Sensors lie unless you validate them (calibration and sanity checks matter)
• Wi-Fi + power are the two biggest hidden costs of "simple" IoT
• Dashboards are only useful if your ingestion pipeline is reliable

Iteration 2 (Concept Only — Not Built)

Iteration 2 was intended to explore:

• Longer-range comms (e.g., LoRa-style links)
• Better power strategy (battery + low-power scheduling)
• More robust ingestion patterns

But I intentionally stopped after Iteration 1 to avoid "architecture cosplay." The first prototype taught me the most important lessons—and proved what needed solving before scaling.

Next Steps (If revisited)

• Ship a second sensor node (prove multi-node ingestion)
• Improve data quality (calibration + outlier handling)
• Add better failure visibility (buffering + retry strategy)