Date: Fri, 29 Mar 2024 12:23:51 +0000 (UTC) Message-ID: <1559264049.267.1711715031385@1a5b09e72cd8> Subject: Exported From Confluence MIME-Version: 1.0 Content-Type: multipart/related; boundary="----=_Part_266_277470874.1711715031384" ------=_Part_266_277470874.1711715031384 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Content-Location: file:///C:/exported.html
Managing Streamed Sensor Data for Mobile Equipment Failure Prediction
<= p>Presented by Prof. Melinda Hodkiewicz and Dr. D=C3=A9bora Corr=C3=AAa=Abstract: The ability to wirelessly stream data from se= nsors on mobile equipment provides opportunities to assess asset condition = proactively.Our streaming data is drawn from a mining industry case study, = containing 23M rows (1.8GB) for a single excavator over nine months. This d= ata has 58 numerical sensor variables and 40 binary indicators describing t= he conditions and status of different subsystems. In addition, data are ava= ilable from the fleet management and maintenance work order systems. We foc= us on the hydraulic subsystem, which has 21 potential failure events report= ed in the period of the data. There are signiffcant issues with the da= ta due to the large volume, inconsistent and asynchronous recording from di= fferent sensors, 57% of rows have missing data, and uncertainties in the gr= ound truth for the dependent variable (hydraulic subsystem failure). We dem= onstrate how the application of an OHLC (Open, High, Low, Close), commonly = used in financial analysis, can be used to compress and manage the data. Se= condly, we create a data frame of OHLC sensor data, fleet management and ma= intenance work order data and demonstrate the application of LASSO penalize= d logistic regression model for variable (sensor/alarm) selection. We = found that the variables selected by the data-driven method have similariti= es when compared to the selection made by experts (asset manager).
= p>