Skip to main content

Edge AI Firmware

The "Small Motor Model" runs entirely on-device. No cloud required for real-time fault detection, efficiency calculation, and predictive insights.

The Small Motor Model

Virtual Dyno

Real-time efficiency calculation without external test equipment.

  • Efficiency: Powerin / Powerout
  • Torque Ripple: Detect mechanical issues
  • Load Profiling: Understand duty cycles
🩺

Machine Doctor

Edge AI inference engine that diagnoses problems before they become failures.

  • 4kHz sampling of vibration + audio
  • FFT analysis buffered in PSRAM
  • Fault flags: Bearing, Imbalance, Alignment
📼

Black Box Recorder

Continuous data logging for post-failure forensics and compliance.

  • MicroSD logging of raw sensor data
  • Timestamped events for incident analysis
  • Exportable data for external tools

User Experience Flow

🔵

Out-of-Box (Desk Mode)

  1. 1.Plug into USB-C power
  2. 2.RGB LED pulses BLUE (Access Point Mode)
  3. 3.Connect via BLE app or browser (192.168.4.1)
  4. 4.Configure Wi-Fi credentials and LoRaWAN keys
🟢

Installation (Field Mode)

  1. 1.Connect high-voltage motor leads
  2. 2.Device auto-detects motor and begins monitoring
  3. 3.LED indicates status at a glance
  4. 4.LoRaWAN heartbeats begin automatically

RGB LED Status Codes

Slow Green Pulse

System Healthy (Motor Running)

Red Flash

Alert (Threshold Exceeded)

Purple Blink

LoRaWAN Packet Sent

Data Strategy

📶 Wi-Fi (Local)

High-bandwidth, real-time data for local monitoring.

  • • WebSocket streaming for live graphs
  • • React dashboard hosted on ESP32-S3
  • • Full sensor resolution (4kHz)
  • • Instant response for troubleshooting

📡 LoRaWAN (Remote)

Low-bandwidth, long-range for fleet monitoring.

  • • Heartbeat every hour (configurable)
  • • Event triggers: "Stall Detected", "Temp Alert"
  • • Works without Wi-Fi infrastructure
  • • Miles of range, years of battery

Plug-and-Play Expansion

Connect any Qwiic/Stemma QT sensor to the expansion port. The firmware automatically scans the I2C bus, identifies the device (OLED screen, temperature sensor, etc.), and integrates it into the dashboard.

🔌Qwiic / Stemma QT (I2C)

Open Source Firmware

The entire firmware stack is open source. Clone the repo, customize the model, and deploy to your fleet.