Mohammad-Khalifah1/Predictive-Maintenance-System ? reverse-engineered prompt

Build me a Streamlit app for predictive maintenance of industrial equipment.

I want a clean dashboard where someone can see equipment health at a glance, upload a CSV of sensor readings, or enter values manually and get instant predictions. The app should estimate remaining useful life, say whether the machine needs maintenance soon, and flag unusual patterns that might look like anomalies. It should also include simple interactive charts so users can explore sensor trends, correlations, and distributions over time, plus a health score that is easy to understand.

Use the sample machine data in the repo as the basis for the experience, and make the app feel polished and easy to use. Include a results page with clear summaries and the option to download the predictions. If you need to check anything about current Streamlit or scikit learn behavior, look up the latest docs online.