• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to previous page

Machine learning study on non-reproducibility of crystal growth results

Research Project

Project/Area Number 18K19033
Research Category

Grant-in-Aid for Challenging Research (Exploratory)

Allocation TypeMulti-year Fund
Review Section Medium-sized Section 30:Applied physics and engineering and related fields
Research InstitutionInstitute of Physical and Chemical Research

Principal Investigator

Kutsukake Kentaro  国立研究開発法人理化学研究所, 革新知能統合研究センター, 研究員 (00463795)

Co-Investigator(Kenkyū-buntansha) 前田 健作  東北大学, 金属材料研究所, 助教 (40634564)
Project Period (FY) 2018-06-29 – 2020-03-31
Project Status Completed (Fiscal Year 2019)
Budget Amount *help
¥6,370,000 (Direct Cost: ¥4,900,000、Indirect Cost: ¥1,470,000)
Fiscal Year 2019: ¥2,210,000 (Direct Cost: ¥1,700,000、Indirect Cost: ¥510,000)
Fiscal Year 2018: ¥4,160,000 (Direct Cost: ¥3,200,000、Indirect Cost: ¥960,000)
Keywords結晶成長 / 機械学習 / 実験再現性 / リアルタイム予測 / 結晶工学 / その場観察 / マルチセンシング / 実験再現 / 応用物理 / プロセス制御 / 再現性 / データ科学
Outline of Final Research Achievements

In crystal growth, the results are often different even though under the same conditions, such as the same a furnace, material, growth recipe, etc. In this research, we aimed to identify the factors of this non-reproducibility using machine learning.
We designed and constructed a multi-sensing system with several sensors in a crystal growth furnace, and collected and managed the data. Furthermore, we succeeded in predicting future temperature changes in real time and quantitatively evaluating the influence of each parameter using a recurrent neural network. These results lead to more accurate crystal growth control.

Academic Significance and Societal Importance of the Research Achievements

本研究は、シミュレーションなどでは考慮することが難しい装置内部の状態の微妙な変化を、実実験でのマルチセンシングによって検出し、その影響を機械学習によって定量化することを目指したものである。実実験におけるデータ取得の指針や時系列データの機械学習の検討など、当初の目的である非再現性の要因追及を越えて、この分野の発展の基礎となる多くの成果が得られた。

Report

(3 results)
  • 2019 Annual Research Report   Final Research Report ( PDF )
  • 2018 Research-status Report
  • Research Products

    (6 results)

All 2019 2018

All Presentation (6 results) (of which Int'l Joint Research: 2 results,  Invited: 4 results)

  • [Presentation] 結晶成長・結晶評価へのデータ科学活用2019

    • Author(s)
      沓掛健太朗
    • Organizer
      第151回結晶工学分科会研究会 いまからはじめるインフォマティクス~チュートリアルから先端事例まで~
    • Related Report
      2019 Annual Research Report
    • Invited
  • [Presentation] Application of data science techniques to crystalline silicon research for solar cells2019

    • Author(s)
      Kentaro Kutsukake
    • Organizer
      29th Workshop on Crystalline Silicon Solar Cells & Modules: Materials and Processes
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research / Invited
  • [Presentation] "AI vs 人間" 良い結晶を創るのはどっち?2019

    • Author(s)
      沓掛健太朗
    • Organizer
      第42回 結晶成長討論会
    • Related Report
      2019 Annual Research Report
    • Invited
  • [Presentation] 機械学習による結晶成長炉内温度の予測2019

    • Author(s)
      沓掛健太朗、前田健作
    • Organizer
      第48回結晶成長国内会議
    • Related Report
      2019 Annual Research Report
  • [Presentation] Generation and propagation of dislocations in multicrystalline silicon for solar cells2019

    • Author(s)
      Kentaro Kutsukake, Yusuke Hayama, Tetsuya Matsumoto, Hiroaki Kudo, Tatsuya Yokoi, Yutaka Ohno, and Noritaka Usami
    • Organizer
      International Symposium on Modeling of Crystal Growth Processes and Devices
    • Related Report
      2018 Research-status Report
    • Int'l Joint Research / Invited
  • [Presentation] データ科学手法による結晶成長炉内の 最適温度測定位置の検討2018

    • Author(s)
      沓掛健太朗、Boucetta Abderahmane、工藤博章、松本哲也、 宇佐美徳隆
    • Organizer
      第47回結晶成長国内会議
    • Related Report
      2018 Research-status Report

URL: 

Published: 2018-07-25   Modified: 2021-02-19  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi