Hard Disk Drives (HDD s) sometimes fail with no apparent reason; some SMART (Self-Monitoring, Analysis and Reporting Technology) attributes present strong correlations with drive failures[3], yet a drive may also fail without (supposedly) any previous indication. Most of the host systems today utilize alert-methods which are reactive by nature-a drive is indicated to fail when some SMART attribute exceeds its vendor defined threshold for valid operation[4]. This approach does not take the cross correlation between different attributes into account and the fact that thresholds vary across different vendors.
No takes yet. Share an insight, caveat, or question.
Gaber et al. (2017) studied this question.
Synapse has enriched 2 closely related papers on similar clinical questions. Consider them for comparative context: