AnasTaherGit/IoT-Plateform-for-Predective-Maintenance — reverse-engineered prompt

Build me a simple IoT predictive maintenance prototype for an industrial machine.

I want a small MicroPython device program that can collect vibration or sensor readings, run an FFT style signal processing step, and send the useful values to a Python side app. The Python app should load a saved machine learning model and use it to predict whether the machine looks normal or may need maintenance soon. I also want a basic Node RED dashboard flow so someone can see the live readings, the prediction result, and a clear warning when the machine is in a risky state.

Keep it practical and easy to run as a prototype, not a huge production system. Include the model loading, sample input data if real sensors are not connected, and clear setup instructions so I can test the full flow locally. If something needs current MicroPython or Node RED docs, look them up online and use the current recommended approach.