Multi-Instrument Scientific Data Normalization Pipeline

Develop a robust data normalization pipeline in Python that can handle time-series data from multiple scientific instruments with varying sampling rates and units. The script should automatically detect data types, handle missing values, perform cross-instrument calibration, and generate standardized DataFrame outputs. Implement advanced statistical techniques like z-score normalization, min-max scaling, and robust scaling, with built-in logging for each transformation step and error handling for heterogeneous data sources.