Key Performance Indicators

KPI	Description	Success Metric
1. Accurate Environmental Monitoring	Thingy:52 node collects temperature, humidity, and air quality (eCO₂) readings.	Sensor values are broadcast via BLE advertising with <5% packet loss and within valid calibration ranges.
2. Reliable BLE Communication	BLE advertising used from Thingy:52 to base node, and from base to actuator node, ensures timely transmission.	Data received every 1–2 seconds with <200ms processing delay on receiving node.
3. Distance Sensing for Presence Detection	Base node uses an ultrasonic sensor to detect proximity events or door state.	>90% detection accuracy with filtered values to avoid noise; stable results under test conditions.
4. Intelligent Vent Control	Base node calculates servo angle using sensor fusion (CO₂ + temperature), sends angle over BLE to actuator node.	Actuator node receives BLE packet and updates vent servo within 2 seconds; smooth transitions with optional manual override.
5. Real-Time MQTT Publishing	Actuator node forwards JSON packets received via BLE to Windows PC, which publishes to Mosquitto broker.	JSON packets printed over UART and published to MQTT within 1 second of BLE receipt.
6. Visualisation and Logging	Grafana dashboard on third PC visualises environmental data via MQTT subscription.	Dashboard reflects live data with ≤2s delay; matches JSON packets; supports cross-verification with CLI.
7. Command Line Interface (CLI)	Shell interface on Ubuntu PC allows control over sensor fusion mode, ultrasonic toggling, and manual servo settings.	Commands execute instantly (<100ms latency); override modes visibly impact system behaviour.
8. BLE Payload Parsing and JSON Formatting	Custom 12-byte BLE advertising packets parsed and converted to JSON on actuator node.	JSON output is correctly structured and printed over UART every cycle with 100% parsing reliability.

Technology Used

Sensors:

• Thingy:52
  • Temperature
  • Humidity
  • Air Quality (eCO₂)
• Base Node (nRF52840DK)
  • Ultrasonic sensor (for proximity/distance sensing)
  • Soil moisture sensor (not implemented in final design)

Actuators:

• Servo motors powered via Adafruit CRICKIT
• Angle set using sensor fusion or CLI override
• M5 Core2 originally considered as a display but replaced by PC dashboard

Wireless Communication:

• Bluetooth Low Energy (BLE)
  • Thingy:52 → Base Node (advertising)
  • Base Node → Actuator Node (advertising)

Processing & Logic:

• Sensor fusion algorithm based on CO₂ + temperature to determine vent angle
• CLI interface to toggle ultrasonic sensing, override sensor values, and switch manual/auto modes
• Manual control of angle and data spoofing for testing

Data Forwarding & Dashboard:

• UART over USB: Base and actuator nodes connected to separate PCs
• Mosquitto MQTT Broker: Hosted on Windows PC connected to actuator node
• Grafana Dashboard: Real-time visualisation of JSON-formatted packets from actuator node