スマートで安心・安全なAI社会を支えるインフラ（半導体検査装置、医用機器、インフラ制御機器、産業機器など）を実現する、先端エレクトロニクス技術とデジタルソリューションを融合した研究開発に取り組んでいます。また、AIを活用した画像処理・認識技術により、半導体をはじめとする製造プロセスの画像検査に関する技術革新と製品実装に取り組んでいます。
電磁両立設計(EMC)、低ノイズ回路技術、高周波/高速伝送、エッジ/AI診断、高信頼/機能安全/ディペンダブルハードウェア、アナログ/デジタル回路、チップレット集積/高電圧/パワエレ実装、機械/深層学習、自動設計、人工知能、コンピュータービジョン、製造検査、画像認識、超解像、予兆診断
Publishing Academic Papers:
[Electro-Magnetic Compatibility: EMC]
U. Paoletti, "Single-Channel Blind Source Separation for Periodic Electromagnetic Switching Noise," in IEEE Transactions on Electromagnetic Compatibility, vol. 65, no. 5, pp. 1300-1308, Oct. 2023, doi: 10.1109/TEMC.2023.3296634.
Abstract:
In this article, a single-channel blind source separation method is proposed to decompose time-domain measurement results of electromagnetic noise into the underlying periodic switching noise source signals without requiring any information about the sources. This is realized by clustering the time-domain waveforms based on their period and similarity. At first, the start and end times of the waveforms need to be defined. Then, periods are determined and a waveform cluster is assigned to each period using a new probabilistic approach. Only in the final step, the waveform shape is considered, when the clusters are filtered by removing the waveform outliers that are too different from the remaining waveforms of each cluster. To this purpose, a distance definition based on the cross correlation among the waveforms is proposed. Using numerical and measurement examples, it is shown that it is possible to determine and rank the contribution to the spectrum of each source separately. This helps individuating the main noise sources more quickly and to analyze the contribution of single sources in a complex noisy environment.
K. Fukumasu, M. Nunokawa, U. Paoletti, K. Matsushima and T. Takami, "Analysis Method for Magnetic Field Strength on On-Board Antenna due to Inverter Common-Mode Noise at Whole Train Level," 2023 International Symposium on Electromagnetic Compatibility – EMC Europe, Krakow, Poland, 2023, pp. 1-6, doi: 10.1109/EMCEurope57790.2023.10274186.
Abstract:
In order to reduce malfunctions of on-board equipment due to noise generated by inverters in rolling stock, a method for analyzing magnetic field noise received by on-board equipment is required. To achieve this, it was necessary to identify the main currents among the noise currents flowing near the onboard antenna and to construct a model that takes into account the phase of those currents. In the work reported here, we identified that the imbalance between the left and right rail currents contributes to the magnetic field noise and constructed a vehicle-level analysis model that includes these noise currents. The analysis confirmed that the model can represent the variation of noise fluctuation for each frame structure around the on-board equipment up to 2 MHz.
[Electronic System Diagnostics]
Hiroaki Itsuji, Takumi Uezono, Tadanobu Toba and Subrata Kumar Kundu, "Real-Time Diagnostic Technique for AI-Enabled System," in IEEE Open Journal of Intelligent Transportation Systems, vol. 5, pp. 483-494, 2024, doi: 10.1109/OJITS.2024.3435712.
Abstract: The last few decades have witnessed a dramatic evolution of Artificial Intelligence (AI) algorithms, represented by Deep Neural Networks (DNNs), resulting in AI-enabled systems being significantly dominant in various fields, including robotics, healthcare, and mobility. AI-enabled systems are currently used even for safety-critical applications, including automated driving, where they encounter reliability challenges from both hardware (HW) and software (SW) perspectives. However, there is no effective technique available that can diagnose HW and SW of AI-enabled systems in real-time during operation. Therefore, this paper proposes an intelligent real-time diagnostic technique for detecting HW and SW anomalies of AI-enabled systems and continuously improving the SW quality during operation. The proposed technique can detect HW anomalies to avoid unexpected changes in AI parameters and subsequent AI performance degradation using single context data with a detection accuracy of more than 92%. The proposed technique can also detect SW anomalies and identify edge cases in real-time, which could result in performance degradation by more than 50% compared to normal conditions. The identified edge cases can be used to continuously enhance the SW quality. Experimental results show the effectiveness of the technique for practical applications and thus can contribute to realize reliable and improved AI-enabled systems.
Hiroaki Itsuji, Takumi Uezono, Tadanobu Toba, Kojiro Ito and Masanori Hashimoto, "A Hardware-Aware Failure-Detection Method for GPU Control-Logic," in IEEE Access, vol. 13, pp. 113890-113904, 2025, doi: 10.1109/ACCESS.2025.3584759.
Abstract: Graphics processing units (GPUs) are used for diverse applications and play a major role even in safety-critical applications. Although performance is usually the primary focus of GPUs, their reliability has become a major concern. One of the undesirable failures in GPUs is silent data corruption (SDC), which causes unexpected outputs without any warning. Various failure detection methods have been proposed for SDCs caused by faults in data units such as registers. However, effective methods for detecting SDCs resulting from faults in control logic, such as scheduling units, have not yet been established. This paper assumes three types of control-logic failures for a general GPU architecture and proposes efficient failure detection methods for each type. For instance, the proposed method efficiently detects GPU-specific control-logic failures caused by program counter faults with a detection rate of 99.5% and can be implemented with a runtime overhead of 5.3% and a memory-resource overhead of 4.2% for a matrix multiplication application. These methods are applicable to a wide range of applications and are expected to enhance system resiliency.
[Analog circuit design]
Wen Li, Masami Makuuchi and Norio Chujo, " Multi-Chips High-Speed and High-Voltage Amplifier," APCCAS2021
Abstract: A linear amplifying circuit with high voltage, high slew rate and wide bandwidth is presented in this paper. The circuit is composed of three chips with different substrate voltage. The maximum output voltage swing can achieve almost twice the withstand voltage. The circuit is verified with a 0.25-um SOI-LDMOS technology with 200V withstand voltage. The experimental result shows that the output voltage swing is 425 Vpp with the slew rate of 1120 V/μsec and bandwidth of 2.3 MHz. The multi-chips amplifier is expected to be used for many industrial applications.
[High speed interconnect]
Yutaka Uematsu; Hideyuki Sakamoto, High Bandwidth and Multi-Channel Power over Coaxial Filters for Automotive Low-Voltage Differential Signaling Interconnect. Proc. ECTC(2020)
Abstract: We propose high bandwidth and low-cost multichannel power over coaxial (PoC) filter configurations for an automotive LVDS interconnect. To enhance the bandwidth of PoC filters toward a higher frequency range, we have designed a small filter pattern on the PoC filter electrodes and demonstrated a filter property improvement of about 5 dB at 10 GHz without degradation of the insertion loss of the signal line in a simulation. We have also designed a multi-channel PoC filter configuration with fewer components and a smaller mounting area and demonstrated sufficient filter performance in a simulation. These structures were fabricated on test boards and demonstrated by measurement.
Norio Chujo, Toru Yazaki, Toshiaki Takai, Takayasu Norimatsu, Yasunobu Matsuoka, and Takashi Takemoto, "Interconnect Technologies for High- Bandwidth ICT Systems," AVID2016
Abstract: The bandwidth of information and communication technologies (ICT) systems, such as servers and routers, doubles every two years. Interconnects thus require higher density, lower loss, and lower power consumption. To meet this requirement, high-speed optical and copper interconnects will be introduced. To realize high-speed interconnects for high-bandwidth ICT equipment, a transceiver for copper cable, an LD driver, and a trans-impedance amplifier (TIA) were developed. Moreover, a small-footprint low-power embedded optical module (EOM), with high-density optical-fiber wiring and an optical connector, was developed.
[Applied Electromagnetics]
Umberto Paoletti; Katsuei Ichikawa; Tomonori Kaneko; Akiko Iizuka
"Contribution of Electric Field Polarization to Temperature Runaway in Radio-Frequency Thawing and Design-Oriented Simulations" in IEEE Access Volume: 12, pp. 157397 - 157407, 21 October 2024, DOI: 10.1109/ACCESS.2024.3484012
Abstract:
In this work the reasons for the occurrence of the temperature runaway effect in radio-frequency thawing are examined and the importance of the polarization of the electric field is clarified. In particular it is observed that the polarization of the electric field normal to the food surface generally counteracts the temperature runaway effect by reducing the food loss in portions at higher temperature. One exception was found to be beef fat. An efficient simulation technique is proposed for the design of food thawing equipment that considers circuit simulations, electromagnetic simulations and thermal calculations. By assuming a uniform food temperature, the simulation time for the whole thawing process is in the order of one second. It is very convenient for the design of the matching circuit and of the overall system, because the variation with time of average temperature, electrode impedance, voltage, efficiency and power balance can be obtained.
[Image processing and recognition technologies utilizing AI and data analytics]
Ito, A., et al., "Super-resolution method for SEM images based on pixelwise weighted loss function", Microscopy 72(5), 2023
https://doi.org/10.1093/jmicro/dfad009
Osaki, M., et al., ″Quantification of three-dimensional pattern-shape variation with CD-SEM top-down image″, Journal of Micro/Nanopatterning, Materials, and Metrology 22(2), 2023
https://doi.org/10.1117/1.JMM.22.2.021008
Fukuda, K., et al, "Trainable die-to-database for large field of view e-beam inspection", Journal of Micro/Nanopatterning, Materials, and Metrology 22(2), 2022.
https://doi.org/10.1117/1.JMM.22.2.021004
[研究トピック]
半導体製造に向けて、10nm以下の微小欠陥を高感度で検出する技術を開発
https://rd.hitachi.co.jp/_ct/17750200
半導体のウェハ接合プロセスにおいて接合部の微小段差を評価する画像処理技術を開発
https://rd.hitachi.co.jp/_ct/17695780
半導体製造プロセスで発生する微小な欠陥の検出精度を高める画像解析技術を開発
https://rd.hitachi.co.jp/_ct/17695278
半導体ウェハ接合部の欠陥を自動検出する超音波検査技術で令和5年度全国発明表彰「日本経済団体連合会会長賞」を受賞
https://rd.hitachi.co.jp/_ct/17686550
[日立評論]
送配電設備のアセットマネジメント最適化に向けた信頼性中心保全（RCM）に基づく保全計画支援技術
(2025/1予定)
LLMを活用したアセットナレッジ自動構築技術
(2025/1予定)
先端半導体デバイスの非破壊検査に対応した新型超音波映像装置FineSAT7
https://www.hitachihyoron.com/jp/papers/2024/03/02/index.html
AI応用高スループット半導体検査ソリューション
https://www.hitachihyoron.com/jp/archive/2020s/2022/01/19/index.html#sec08
ウエーハ全面検査を実現する深層学習ベース半導体検査技術
https://www.hitachihyoron.com/jp/archive/2020s/2021/01/24/index.html#sec04
2015年 第15回（平成27年度）山崎貞一賞 半導体及び半導体装置分野
2021年度 第70回 電機工業技術功績者表彰
ものづくり部門 優秀賞(2021) 「パワー半導体製品の高信頼化を実現するはんだ接合技術」
2022年度(第71回) 電機工業技術功績者表彰
重電部門 最優秀賞(2022) 「800V駆動電気自動車用インバータのノーマルモードノイズ抑制技術」
第66回電気科学技術奨励賞 文部科学大臣賞(2018)
「最先端半導体微細パターンの高精度計測を実現する走査型電子顕微鏡の回路ノイズ相殺と帯電抑制技術の開発」
第68回電気科学技術奨励賞 電気科学技術奨励会会長賞(2020)
「世界初の 800V 駆動電気自動車を実現するインバータ低ノイズ化技術」
第71回電気科学技術奨励賞 (2023)
「高速道路での手放し運転を可能とする高度先進運転支援 ECU を片手サイズで実現するハードウェア設計技術」
12th IEEE CPMT Symposium Japan Young Researcher Award 2023
"Abnormal Position Identification Technique for Communication Channel using Adaptive DFE"
溶接学会 マイクロ接合研究委員会 マイクロ接合優秀研究賞 (2025)
「長期保存電⼦部品のはんだ濡れ性改善に関するギ酸処理再生技術」
日立パワーソリューションズが、デバイス内面に発生する剥離・ボイドなど欠陥検出画像の高画質化と300mmウェーハの一括検査を可能にした超音波映像装置を発売
https://www.hitachi.co.jp/New/cnews/month/2022/08/0829.html
EUV適用で高まる検査計測ニーズに対応した電子線広視野検査システム「GS1000」を開発
https://www.hitachi.co.jp/New/cnews/month/2021/12/1213.html
通信装置のソフトエラー対策、ITU-T国際標準制定
https://www.hitachi.co.jp/New/cnews/month/2018/11/1122a.html
