KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 10 of 25

Some of the metrics are blocked by your
consent settings
PSO-Optimized Deep Learning for Ultra-Precise Corrosion Detection on HDD Read/Write Heads
(2025-01-01)
Punyammaree, Chaiwat
;
Kaitwanidvilai, Somyot
This paper presents a novel deep learning approach for automated detection and counting of corrosion pits on Hard Disk Drive (HDD) read/write heads using Scanning Electron Microscopy (SEM) images. A U-Net model optimized via Particle Swarm Optimization (PSO) is developed to enhance segmentation performance by automatically tuning hyperparameters. The methodology includes optimized SEM image acquisition, preprocessing (patch-based subdivision and expert annotation), PSO-driven hyperparameter selection, and post-processing with thresholding and connected component analysis for pit counting. Experimental results demonstrate that the PSO-optimized U-Net significantly outperforms standard U-Net, SegNet, and LinkNet models, achieving an F1-score of 79.60%, an IoU of 86.51%, and an accuracy of 99.77%. Additionally, the proposed method achieves 86.9% counting accuracy, surpassing human experts (72.7%) while processing images 15 times faster (180 seconds vs. 2700 seconds per image). These findings highlight the potential of PSO-optimized deep learning for improving HDD quality control by providing an accurate, efficient, and standardized solution for corrosion pit detection, ultimately reducing the risk of HDD failure and data loss.
Some of the metrics are blocked by your
consent settings
OPTIMIZATION OF PROCESS PARAMETERS IN LASER TRANSMISSION BONDING TO INCREASE THE SHEAR STRENGTH OF AU-SN (IMCS) JOINT
(2024-01-01)
Khunjun, Narttakarn
;
Sirivongpaisal, Nikorn
;
Deeying, Jakkawat
;
Limsuwan, Pichet
Abstract Au-Sn solder has been currently used for soldering the laser diode silicon chip on the slider surface which is alumina (Al<inf>2</inf>O<inf>3</inf>) titanium carbide (TiC) in hard disk drive (HDD). However, intermetallic compounds (IMCs) form at the interface during the bonding process and result in cracks within these interfacial IMCs. As a consequence, the bonded joint failure occurs for Au-Sn (IMCs) joints with low shear stress. The shear stress is considered a major contributor to the quality and reliability of HDD. At present, the average shear strength of the Au-Sn (IMCs) joint is approximately 2.716 MPa. In this work, 4,572 Au-Sn bonded samples using the laser transmission bonding process were analysed to find suitable bonding factors to improve the shear stress of Au-Sn (IMCs). The analyse phase utilizes heat map correlation to determine 5 key process input variables (KPIVs) including (1) Temperature, (2) Stage Cross-Track (CT), (3) Stage Down-Track (DT), (4) Z-height (5) Laser power. The key process output variable (KPOV) is the shear strength of the IMCs layer between the laser diode silicon chip and the Alumina Titanium Carbide (AlTiC) substrate. Then, a response surface methodology (RSM) is performed to provide the significant factors that generate a reliable mathematical relationship between the process parameters and the desired response. The confirmation experiment performed illustrates that the new process setting of the IMCs shear strength layer is increased from 2.716 MPa to 5.050 MPa which accounts for 85.94% which is a significant improvement.
Some of the metrics are blocked by your
consent settings
A deep learning system for recognizing and recovering contaminated slider serial numbers in hard disk manufacturing processes
(2021-09-01)
Chousangsuntorn, Chousak
;
Tongloy, Teerawat
;
Chuwongin, Santhad
;
Boonsang, Siridech
This paper outlines a system for detecting printing errors and misidentifications on hard disk drive sliders, which may contribute to shipping tracking problems and incorrect product delivery to end users. A deep-learning-based technique is proposed for determining the printed identity of a slider serial number from images captured by a digital camera. Our approach starts with image preprocessing methods that deal with differences in lighting and printing positions and then progresses to deep learning character detection based on the You-Only-Look-Once (YOLO) v4 algorithm and finally character classification. For character classification, four convolutional neural networks (CNN) were compared for accuracy and effectiveness: DarkNet-19, EfficientNet-B0, ResNet-50, and DenseNet-201. Experimenting on almost 15,000 photographs yielded accuracy greater than 99% on four CNN networks, proving the feasibility of the proposed technique. The EfficientNet-B0 network outperformed highly qualified human readers with the best recovery rate (98.4%) and fastest inference time (256.91 ms).
Some of the metrics are blocked by your
consent settings
Simple Soft-Information Developers Based on Parallel Detection Scheme in Bit-Patterned Media Recording (BPMR) Systems
(2021-06-27)
Mattayakan, Mutita
;
Sokjabok, Siwakon
;
Warisam, Chanon
Among the alternate ultra-high density data storage technologies, bit-patterned media recording (BPMR) is one of the mentioned technologies that was proposed to achieve a higher density. Due to the closeness between the magnetic islands, so the readback sequences are disrupted from the effect of two-dimensional (2D) interference, which includes both intersymbol interference (ISI) and intertrack interference (ITI). In this paper; therefore, we propose the simple and effective soft-information developers (SIDs) to improve the soft-information that are obtained from both vertical and horizontal detectors. An optimal weight was optimized under the BPMR channel before multiplying with the horizontal soft-information. The proposed SIDs and conventional 2D detector's performances were carefully compared in the terms of BER. The results show that our proposed methods achieve a gain of approximately 8 dB at a BER of 10-3.
Some of the metrics are blocked by your
consent settings
An Intertrack Interference (ITI) subtraction scheme for bit-patterned media recording
(2021-05-19)
Buajong, Chaiwat
;
Warisarn, Chanon
;
Koonkarnkhai, Santi
;
Kovintavewat, Piya
Currently, hard disk drive has been struggled to overcome the super-paramagnetic limit that restricts the density increment. Bit-patterned magnetic recording (BPMR) is a candidate that can increase an areal density (AD) up to 4 Tb/in2. However, inter-track interference (ITI) arising from a narrow track width at high AD severely degrades the system performance. This study proposes the ITI subtraction technique with turbo iteration in a coded BPMR system. This method refines the equalized sequence by subtracting the ITI using an imitated ITI sequence that is generated by the soft information obtained from turbo iteration and ITI coefficients. The refined sequence that contains the partial ITI is sent to the turbo iteration as many rounds as needed. Simulation results indicate that the proposed system outperforms the conventional system whether or not media noise is considered.
Some of the metrics are blocked by your
consent settings
Safety path planning with obstacle avoidance using particle swarm optimization for agv in manufacturing layout
(2019-02-01)
Praserttaweelap, Rawinun
;
Kaitwanidvilai, Somyot
;
Aoyama, Hisayuki
In robotic systems, path planning is the one of important processes for robot motion. The best path planning is required for shortest path searching that can make fast movement of robot. However, the real environment is not only the path from point to point but it has obstacles which are the one of constraints for best path searching. The obstacle avoidance is concerned to avoid the crashing between robot and obstacle under environment. In Hard Disk Drive manufacturing, the first priority is safety constraint for non-collision and second priority is shortest path for processing time saving. This research designed the algorithm for path planning and obstacle avoidance for AGV in Hard Disk Drive Manufacturing of Seagate Technology (Thailand) Ltd by using particle swarm optimization. The fitness function on particle swarm optimization process for particle searching has been integrated with obstacle avoidance function to find the best path for robot without collision and total distance to find the shortest path. This algorithm is applied to verifying the model performance. The simulation results of this research are done by MATLAB 2016b and illustrate the good performance on different cases with controlled parameter.
Some of the metrics are blocked by your
consent settings
Hybrid adaptive notch filter and fixed-structure pid H ∞ robust loop shaping control based pso for hard disk drive servo actuator
(2019-02-01)
Konghuayrob, Poom
;
Kaitwanidvilai, Somyot
;
Aoyama, Hisayuki
In order to achieve the high precision head-positioning of the voice coil motor (VCM) actuator with narrow track pitch, the adaptive notch filter based limited-search-range of particle swarm optimization (PSO), as well as the fixed-structure propor-tional-integral-derivative H <inf>∞</inf> robust loop shaping controller using the concept of four closed loop disturbance norms is proposed. Generally the conventional method, fixed-frequency notch filter (FFNF), is combined with the nominal plant to reduce the effect of the mechanical vibration resonance; however, the resonance mode of servo system can be shifted with various factors such as the ambient temperature change, and the unbalanced disk. In addition, mathematical solving in the H <inf>∞</inf> robust control problems and the suitable notch filter design are very complex and the final results of the conventional controller with notch filters are normally complicated structure and high order which is difficult to implement. Thus, the adding of intelligent system in the proposed design with the careful range of the search space is utilized to suppress the vibration caused by resonance mode shifting and also reduce the order of the final robust controller. Simulation results of six scenarios test demonstrate the effectiveness of the proposed design compared with FFNF in the commercial product.
Some of the metrics are blocked by your
consent settings
Read-reference BER categorization improvement in WPOH measurement of HAMR drive
(2018-07-02)
Jariampan, Panadda
;
Warisarn, Chanon
Bit-error rate (BER) is used as a primary metric in write power-on hours (wPOH) measurement of heat-assisted magnetic recording (HAMR) drive. BER is measured after writing process represented writing quality while read-only BER showed reading quality or read reference in the measurement. The read-reference BER usually remains close to its initial value, however, in some cases it shows BER fluctuation or degradation as well. Therefore, in this paper focuses on the read-reference BER categorization and an improvement of the categorization algorithm.
Some of the metrics are blocked by your
consent settings
The investigation of corrosion behavior of electrodeposited Co-Fe alloys surface with phase and 3D reconstruction of image using TIE and TPE
(2018-01-01)
Srisuwan, Thanthanat
;
Somphonsane, Ratchanok
;
Sutthiruangwong, Sutha
;
Buranasiri, Prathan
In this research, phase retrieval using transport of intensity equation (TIE) and transport of phase equation (TPE) have been used for determine the surface of corrosion behavior of electrodeposited Co-Fe alloys, an important component in hard disk drive. Without interferometer as used in digital holography technique, by using TIE and TPE, our observing setup is simple and compact. The experimental results show the capability of TIE and TPE for observing the pitting corrosion. The pitting corrosion of Co-Fe image results have also shown. The dynamic process of microstructure surface system would be conducted and may be applied to hard disk drive industry in the future.
Some of the metrics are blocked by your
consent settings
Low order robust ν-gap metric H∞ loop shaping controller synthesis based on particle swarm optimization
(2017-12-01)
Konghuayrob, Poom
;
Kaitwanidvilai, Somyot
The strong demand of the data storage capacity has been increasing significantly. According to the heat-assisted magnetic recording (HAMR) technology, the demand trend of hard disk drive (HDD) is predicted that the areal density will be achieved 10 Tbit/in<sup>2</sup> before the year 2020. High areal density results in a narrow track pitch which is quite sensitive to the external disturbance including the measured noise. This point is the benchmark problem of the high precision controller design for controlling the HDD servo mechanism. Moreover, the systematic uncertainties have to be taken into consideration in controller design procedure as well. The alternative robust ν-gap metric related to H<inf>∞</inf> loop shaping is proposed in this paper to stabilize a voice coil motor in HDD under the uncertainty condition. The potential particle swarm optimization (PSO) is adopted to minimize the gap between the plant with H<inf>∞</inf> controller and the plant with specified 3 controller orders. Instead of using the conventional H<inf>∞</inf> controller with high order with a complicated structure, this paper applies the proposed lower controller order based on ν-gap which is more appropriate implement in the actual application. The performance and robustness of both controllers are compared in the simulation studies. The results confirm the similar characteristics of both controllers in terms of performance tracking and disturbance rejection. Furthermore, the system stability index called stability margin with 0.472 and system perturbations testing condition also emphasizes the robustness and effectiveness of the proposed controller.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results