US 12,066,908 B2
System and method for predicting and avoiding hardware failures using classification supervised machine learning
Deeder M. Aurongzeb, Austin, TX (US); Malathi Ramakrishnan, Madurai (IN); and Parminder Singh Sethi, Punjab (IN)
Assigned to DELL PRODUCTS LP, Round Rock, TX (US)
Filed by Dell Products, LP, Round Rock, TX (US)
Filed on Jul. 29, 2022, as Appl. No. 17/877,076.
Prior Publication US 2024/0036999 A1, Feb. 1, 2024
Int. Cl. G06F 11/20 (2006.01); G06F 9/50 (2006.01)
CPC G06F 11/2033 (2013.01) [G06F 9/5044 (2013.01); G06F 9/5055 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A hardware failure prediction and avoidance system executing on a unified endpoint management (UEM) platform information handling system comprising:
a network interface device to receive failed operational telemetries for a plurality of client information handling systems including power analytics, software application analytics, and event viewer error logs, wherein each failed operational telemetry includes an error associated with a labeled hardware type;
a processor executing code instructions of the hardware failure prediction and avoidance system to:
execute a classification supervised learning algorithm on the failed operational telemetries to determine a failure probability, meeting a preset failure probability threshold value, that an adjustable hardware or software configuration described by future operational telemetries will co-occur with the error of the labeled hardware type;
identify the adjustable hardware or software configuration as a problematic system configuration;
the network interface device to receive a current operational telemetry for a first client information handling system including current power analytics, current software application analytics, and current event viewer error logs;
the processor to identify the problematic system configuration by correlating the current operational telemetry with a portion of the failed operational telemetries such that the failure probability meets the preset failure probability threshold value; and
the network interface device to transmit a recommendation to the first client information handling system to adjust or terminate the problematic system configuration in order to avoid occurrence of the error identified in the failed operational telemetries at the first client information handling system.