Continuous-time digital twin with analog memristive neural ordinary differential equation solver
Digital twins, which replicate real-world entities through computational models, are transforming manufacturing and automation. While recent advances in machine learning have enabled data-driven digital twin development using discrete-time data and finite-depth models on digital hardware, these approaches face significant limitations. They struggle to capture continuous-time dynamics and model complex systems, and suffer from substantial time and energy overheads due to physically separated storage and processing as well as frequent analog-digital (A/D) conversions. Here, we propose a memristive neural ordinary differential equation (ODE) solver for digital twins. Our approach is intrinsically time-continuous using infinite-depth neural networks to model complex dynamics. Fully analog memristor arrays collocate storage and computation, addressing the von Neumann bottleneck and reducing A/D conversion requirements. We experimentally validate our solver on digital twins of HP variable-resistor model and Lorenz96 dynamics, demonstrating a 166.5-fold/369.3-fold speedup and a 499.0-fold/673.9-fold improvement in energy efficiency, respectively. This work paves the way to future digital twins for Industry 4.0.
Authors
- Dashan Shang (ORCID: https://orcid.org/0000-0003-3573-8390)
- Meng Xu (ORCID: https://orcid.org/0000-0003-1674-6037)
- Songqi Wang
- Jia Chen (ORCID: https://orcid.org/0000-0003-4814-5420)
- Hegan Chen (ORCID: https://orcid.org/0000-0002-9661-2727)
- Xiaoshan Wu (ORCID: https://orcid.org/0000-0002-7003-0884)
- Dingchen Wang (ORCID: https://orcid.org/0000-0003-3747-1626)
- Yi Li (ORCID: https://orcid.org/0000-0001-8061-6099)
- Kwang‐Ting Cheng (ORCID: https://orcid.org/0000-0002-3885-4912)
- Qingsheng Zhu (ORCID: https://orcid.org/0000-0002-0643-4526)
- Shaocong Wang (ORCID: https://orcid.org/0000-0002-7195-8676)
- Jichang Yang (ORCID: https://orcid.org/0000-0003-3760-6762)
- Xinyu Tian (ORCID: https://orcid.org/0009-0009-1522-5951)
- Zhongrui Wang (ORCID: https://orcid.org/0000-0003-2264-0677)
- Yangu He (ORCID: https://orcid.org/0009-0008-9035-1686)
- Xi Chen (ORCID: https://orcid.org/0000-0003-4221-4288)
- Ming Liu (ORCID: https://orcid.org/0000-0002-0937-7547)
- Xinyuan Zhang (ORCID: https://orcid.org/0000-0002-2974-8392)
- Yi Li (ORCID: https://orcid.org/0000-0003-1930-8550)
- Han Wang (ORCID: https://orcid.org/0000-0001-5121-3362)
- Qi Liu (ORCID: https://orcid.org/0000-0001-7062-831X)
- Ning Lin (ORCID: https://orcid.org/0000-0002-6198-7043)
- Xiaojuan Qi (ORCID: https://orcid.org/0000-0002-4285-1626)
- Yifei Yu (ORCID: https://orcid.org/0009-0003-9208-2903)
- Xumeng Zhang (ORCID: https://orcid.org/0000-0002-3828-151X)
- Yinan Lin
Institutions
- Hong Kong University of Science and Technology (HK)
- Huazhong Agricultural University (CN)
- Fudan University (CN)
- Southern University of Science and Technology (CN)
- Institute of Microelectronics (SG)
- Integrated Sensors (United States) (US)
- Institute of Microelectronics (CN)
- University of Chinese Academy of Sciences (CN)
- Huazhong University of Science and Technology (CN)
- University of Hong Kong (HK)
Publication Details
- Journal
- Science Advances
- Published
- 2025-05-28
- DOI
- https://doi.org/10.1126/sciadv.adr7571
- Citations
- 3
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- article
- Field-Weighted Citation Impact
- 2.19