On May 29, 2025, ULINK updated its DA Drive Analyzer AI model for predicting drive failures in over 99% of commercial drives, including SATA, NVMe, and SCSI drives. The new AI model boasts improved performance scores for SATA drives over the previous AI model as well as SMART Trips.
An AI-based Drive Failure Prediction Service
For those new to DA Drive Analyzer, it is software that monitors the health of your computer storage drives and warns you when your drives are in danger of failing. At its core is a machine learning model that detects subtle patterns in drive health data to predict upcoming drive failures. The software is currently offered for multiple platforms including QNAP NAS, ASUSTOR NAS, and Windows PC. It works by uploading non-sensitive drive health data to a secure cloud server where the data is analyzed for health issues. Alerts can be issued via email, to an application that runs on the platform, or to an online dashboard that you can log into from anywhere. The PC version of this software (which is called DA SmartQuest) also offers optional on-device only AI-based drive failure predictions for those with high network security needs.
Improved 180-Day Precision
With the latest AI model, the 180-day precision goes up from 68% in the last model to 83% in the current model for SATA HDDs, and from 43% in the last model to 78% in the current model for SATA SSDs. What this means is that whenever the last model predicted 100 SATA HDDs and 100 SATA SSD would fail, 68 SATA HDDs and 43 SATA SSDs would actually fail over the next 180 days. Now, if the AI model predicts these same drives will fail, 83 SATA HDDs and 78 SATA SSDs will fail over the next 180 days. Having precision go up with the latest model means that there will be fewer false alarms, and fewer wasted drive replacements.
In contrast, SMART Trips only offer a 180-day precision of 52% for SATA HDDs and a 47% precision for SATA SSDs. Having a lower precision means that a SMART trip is more likely to result in a false alarm and waste drive replacements than DA Drive Analyzer.
Overall, a higher precision translates into less wasted money buying replacement drives, and fewer wasted hours trying to rebuild RAID.
Higher Recall Rates
The latest AI model also delivers a 2.3x higher recall rate for SATA HDDs over the last AI model. This means that for every failing drive that the last model caught, the latest model will catch 2 or more. A higher recall means that fewer dying drives go undetected.
The latest AI model also delivers a 13.7x higher recall rate for SATA HDDs over SMART trips. Having a lower recall rate means that SMART trips will miss a lot of the failing drives that DA Drive Analyzer can catch.
Overall, a higher recall rate translates into fewer undetected failed drives and less data lost due to unforeseen drive failures.
“S.M.A.R.T. alone isn’t enough—many users are caught off guard by silent drive failures,” said Tim Lin, Product Manager at QNAP. “DA Drive Analyzer uses an AI model trained on behavioral data from millions of drives. It can detect potential anomalies even before any S.M.A.R.T. warning is triggered, enabling users to take preventive action and avoid data loss.”
“DA Drive Analyzer goes beyond traditional drive failure diagnostics, demonstrating ULINK’s leadership and innovation in AI and storage technology,” said Joseph Chen, CEO of ULINK Technology.
DA Drive Analyzer is constantly improving. As we collect more drive health data, we are better able to optimize the performance of our AI models for detecting drive failures. Find out how you can try DA Drive Analyzer today.
See QNAP’s press release.
Recent Comments