NASA Milling Dataset

Dataset History

The data in this set represents experiments from runs on a milling machine under various operating conditions. In particular, tool wear was investigated (Goebel, 1996) in a regular cut as well as entry cut and exit cut. Data sampled by three different types of sensors (acoustic emission sensor, vibration sensor, current sensor) were acquired at several positions.

There are 16 cases with varying number of runs. The number of runs was dependent on the degree of flank wear that was measured between runs at irregular intervals up to a wear limit (and sometimes beyond). Flank wear was not always measured and at times when no measurements were taken, no entry was made.

Dataset Description

The data is organized in a 1x167 matlab struct array with fields as shown.

Field name - Description

• case - Case number (1-16)
• run - Counter for experimental runs in each case
• VB - Flank wear, measured after runs; Measurements for VB were not taken after each run
• time - Duration of experiment (restarts for each case)
• DOC - Depth of cut (does not vary for each case)
• feed - Feed (does not vary for each case)
• material - Material (does not vary for each case)
• smcAC - AC spindle motor current
• smcDC - DC spindle motor current
• vib_table - Table vibration
• vib_spindle - Spindle vibration
• AE_table - Acoustic emission at table
• AE_spindle - Acoustic emission at spindle

There are 16 cases with varying number of runs. The number of runs was dependent on the degree of flank wear that was measured between runs at irregular intervals up to a wear limit (and sometimes beyond). Flank wear was not always measured and at times when no measurements were taken, no entry was made.

The dataset is also available in CSV format which has been converted (mat to csv) using this Python Code.

Accessing the Dataset

We have made this dataset available on Kaggle. Watch out for Official NASA Website.

The dataset is in .mat format and also contain brief overview of documentation (README.pdf) by the authors itself.

The dataset is also available in CSV format which has been converted (mat to csv) using this Python Code.

Acknowledgment

A. Agogino and K. Goebel (2007). BEST lab, UC Berkeley. "Milling Data Set ", NASA Ames Prognostics Data Repository (http://ti.arc.nasa.gov/project/prognostic-data-repository), NASA Ames Research Center, Moffett Field, CA

---