I need a complete edge-AI solution that lets a Raspberry Pi watch over a 0.5 kW AC motor and cut the power the moment an anomaly appears. The Pi will gather vibration, temperature, and speed data, run an on-device anomaly-detection model, and trigger a high-power relay (or contactor) through its GPIO pins to stop the motor when necessary. Hardware & sourcing Every part—Pi, sensors, relay, wiring, and any signal-conditioning boards—has to be locally available from uruktech.com, ardunic.com, or ecity-iq.com. I don’t have brand preferences, so feel free to choose components that meet accuracy and safety requirements. AI & firmware I’m open to either an Edge Impulse workflow or a pure-Python/TensorFlow-Lite pipeline; choose whichever lets you reach real-time performance on the Pi. The key is reliable anomaly detection on raw or pre-processed vibration, temperature, and speed signals, with minimal false alarms. What matters most • Clean Python code, well commented • Safe relay wiring and motor-rated protections (remember: 0.5 kW, 220 V) • A repeatable training/inference setup—dataset preparation, model training, and deployment steps clearly documented Please type: "I've read the full description in your quote" • Automatic motor shutdown tied to the model’s anomaly score threshold (configurable) • A short user guide so I can reproduce everything from wiring to model retraining on another Pi Acceptance The project is done when the system can run for several hours, detect an injected fault (e.g., weighted unbalance), and disengage the motor within two seconds, while logging the event to a CSV file. If you’re comfortable with Python, signal processing, and AC safety, let’s get this motor smart and safe.