Dynamic Mass Spectrometry Data Pipeline with Error Handling

Design a robust Python pipeline for processing large-scale mass spectrometry data from multiple scientific instruments. The script must handle variable file formats (mzXML, .raw, .mgf), implement automatic data normalization using numpy, and create a fault-tolerant processing mechanism that logs errors without interrupting entire batch processing. Include type hints, implement comprehensive logging, and ensure the pipeline can process files from different mass spec vendors like Thermo, Bruker, and Waters.