HDD magnetic head degradation field-failure detection and prediction

Patent metadata

Field	Value
Publication number	US-9548070-B1
Application number	US-201514750815-A
Country	US
Kind code	B1
Filing date	Jun 25, 2015
Priority date	Jun 25, 2015
Publication date	Jan 17, 2017
Grant date	Jan 17, 2017

Abstract

A method, apparatus, and system for implementing head degradation detection and prediction for hard disk drives (HDDs) are provided. Statistical parameter stability tests are used that are calculated as Confidence Intervals (CIs) from HDD parameter measurements. HDD parameters include, for example, change or delta in fly height (dFH), magnetic resistor resistance (MRR) head values, servo variable gain amplifier (SVGA) readback amplitude values, that are used in conjunction with two or more confidence level indicators and an Auto-Regressive Integrated Moving-Average (ARIMA) (p, d, q) predictor for head degradation detection and field failure prediction.

First claim

What is claimed is:

1. A method for implementing head degradation detection and field failure prediction for hard disk drives (HDDs), said method comprising:
calculating two or more confidence level indicators for predefined HDD head parameters;
comparing said calculated confidence level indicators with a respective predefined threshold value for the predefined HDD head parameters for providing a head failure prediction; and identifying both soft error rate (SER) fluctuations, and an error recovery procedure (ERP) count exceeding a prespecified threshold, to provide the head failure prediction.

2. The method as recited in claim 1 , wherein said predefined HDD head parameters include change or delta in fly height (dFH), magnetic resistor resistance (MRR) value of the head, and servo variable gain amplifier (SVGA) readback amplitude values.

3. A method for implementing head degradation detection and field failure prediction for hard disk drives (HDDs), said method comprising:
calculating two or more confidence level indicators for predefined HDD head parameters including calculating confidence intervals (CI) using maximum likelihood estimate (MLE) for predefined HDD head parameters; and
comparing said calculated confidence level indicators with a respective predefined threshold value for the predefined HDD head parameters for providing a head failure prediction.

4. The method as recited in claim 3 , further includes calculating confidence intervals (CI_bootstrap) using moving block bootstrap (MBB) for predefined HDD head parameters.

5. The method as recited in claim 4 , includes calculating difference confidence interval values of the confidence intervals (CI) and (CI_bootstrap) for predefined HDD head parameters.

6. The method as recited in claim 5 , wherein comparing said calculated confidence level indicators with a respective predefined threshold value for the predefined HDD head parameters includes comparing said calculated difference confidence interval values with respective predefined threshold value for the predefined HDD head parameters including one or more of change or delta in fly height (dFH), magnetic resistor resistance (MRR) value of the head, and servo variable gain amplifier (SVGA) readback amplitude values.

7. An apparatus for implementing head degradation detection and prediction for hard disk drives (HDDs), comprising:
a controller;
said controller calculating two or more confidence level indicators for predefined HDD head parameters;
said controller comparing said calculated confidence level indicators with a respective predefined threshold value for the predefined HDD head parameters for providing a head failure prediction; and
said controller identifying both soft error rate (SER) fluctuations and an error recovery procedure (ERP) count exceeding a prespecified threshold to provide the head degradation prediction.

8. The apparatus as recited in claim 7 includes control code stored on a non-transitory computer readable medium, and wherein said controller uses said control code to implement head degradation detection.

9. The apparatus as recited in claim 7 wherein said predefined HDD head parameters include change or delta in fly height (dFH), magnetic resistor resistance (MRR) value of the head, and servo variable gain amplifier (SVGA) readback amplitude values.

10. The apparatus as recited in claim 7 wherein said controller calculating two or more confidence level indicators for predefined HDD head parameters includes said controller calculating confidence intervals (CI) using maximum likelihood estimate (MLE) for predefined HDD head parameters.

11. The apparatus as recited in claim 10 includes said controller L-Lead time head parameter value predictions using an Auto Regressive Integrated Moving Average (ARIMA) HDD parameter time-series predictor apparatus.

12. The apparatus as recited in claim 11 wherein said Auto Regressive Integrated Moving Average (ARIMA) HDD parameter time-series predictor apparatus includes moving block bootstrap (MBB) for predefined HDD head parameters and ARIMA (p, d, q) filter coupled to said moving block bootstrap (MBB), wherein said moving block bootstrap (MBB) includes a bootstrap (block scrambled averaging) function, and said ARIMA(p, d, q) filter represent a finite-difference equation defined by AR-polynomial φ(B), of degree p, MA-polynomial θ(B) of degree q, in the variable B, BX T =X T−1 is the backward shift operator
θ( B )(1− B ) d X t =θ( B ) X* T , where

(1−B) d is the d-th differencing-operator that transforms {X T } into a stationary time series; and {X* T } is its moving-block-average.

13. The apparatus as recited in claim 11 wherein said ARIMA (p, d, q) filter receives AR polynomial coefficients and MA polynomial coefficients, and said ARIMA (p, d, q) filter receiving inputs X* T PN time series and X T d-differencing—data time series from said bootstrap (block scrambled averaging) function, and providing maximum likelihood estimate output X* T+L , where L=lead time predictor.

14. The apparatus as recited in claim 11 includes said controller calculating difference confidence interval values of the confidence intervals (CI) and (CI_bootstrap) for predefined HDD head parameters, and wherein said controller comparing said calculated confidence level indicators with a respective predefined threshold value for the predefined HDD head parameters includes said controller comparing said calculated difference confidence interval values with a respective predefined threshold value for the predefined HDD head parameters.

15. A system for implementing head degradation detection and prediction for hard disk drives (HDDs), comprising:
a hard disk drive (HDD), said HDD comprising a controller;
at least one disk; said disk including a disk media for storing data;
a slider carrying a head for reading and writing data;
said controller calculating two or more confidence level indicators for predefined HDD head parameters; and
said controller comparing said calculated confidence level indicators with a respective predefined threshold value for the predefined HDD head parameters including Auto Regressive Integrated Moving Average (ARIMA)(p, d, q)-prediction for providing head degradation detection and field failure prediction.

16. The system as recited in claim 15 includes control code stored on a non-transitory computer readable medium, and wherein said controller uses said control code to implement head degradation detection.

17. The system as recited in claim 15 wherein said predefined HDD head parameters include change or delta in fly height (dFH), magnetic resistor resistance (MRR) value of the head, and servo variable gain amplifier (SVGA) readback amplitude values used in an ARIMA(p,d,q)-field failure predictor.

18. The system as recited in claim 15 wherein said controller calculating two or more confidence level indicators for predefined HDD head parameters includes said controller calculating confidence intervals (CI) using maximum likelihood estimate (MLE) for predefined HDD head parameters, and said controller calculating confidence intervals (CI_bootstrap) using moving block bootstrap (MBB) for predefined HDD head parameters used in an ARIMA(p,d,q)-field failure predictor.

19. The system as recited in claim 18 includes said controller includes said controller calculating difference confidence interval values of the confidence intervals (CI) and (CI_bootstrap) for predefined HDD head parameters, and wherein said controller comparing said calculated confide

Assignees

• HGST Netherlands BV
• Western Digital Tech Inc

Inventors

• Chatradhi Sridhar
• Hassner Martin Aureliano
• Letourneaut Natasa
• Yamamoto Satoshi

Classifications

• G11B5/455 Primary

  Arrangements for functional testing of heads; Measuring arrangements for heads · CPC title

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06N7/005

  Physics · mapped topic