17–22 May 2026
C.I.D
Europe/Zurich timezone

Leveraging low-cost sensors and deep autoencoders for pump health monitoring in accelerator facilities

MOP6303
18 May 2026, 16:00
2h
C.I.D

C.I.D

Deauville, France
Poster Presentation MC6.D13: Instrumentation: Artificial Intelligence Poster session

Speaker

Joseph Calvey (Argonne National Laboratory)

Description

The reliability of cooling systems is critical for removing substantial amounts (in megawatts) of waste heat from numerous high-power accelerator components (e.g., magnets, RF structures, power supplies) and beamline components*. Reliance on manual inspection of hundreds of pumps is inefficient and increases the risk of costly component damage and unplanned downtime. This study introduces a real-time method for automated vibration monitoring of pumps designed to help transition facility operations from reactive to predictive maintenance. Our approach integrates (i) affordable vibration sensors** and (ii) Deep Autoencoder*** models for detecting anomalies. We have deployed vibration sensors on Bunch Lengthening System (BLS) Helium and water pumps, where Linux nodes aggregate and preprocess the hourly collected data. To support operational decision-making, a web-based diagnostic platform is being developed, offering real-time visualization of vibration trends against weekly baseline data and generating Short-Time Fourier Transform (STFT) spectrograms for detailed frequency analysis. Additionally, the web platform will show hourly inferences from the Deep Autoencoder on the data stream, autonomously detecting spectral anomalies indicative of mechanical faults. The integration of scalable edge data collection, advanced visualization, and unsupervised deep learning will provide a vital safeguard for maintaining operational readiness in particle accelerators.

Footnotes

E. Swetin, M. Kirshenbaum, and C. Putnam, “Cooling water systems for accelerator components at the Advanced Photon Source”, in Proc. MEDSI’02, Argonne, IL, USA, Sep. 2002.
STMicroelectronics, “IIS3DWB: Ultra-Wide-Bandwidth, Low-Noise, 3-Axis Digital Vibration Sensor”, DS13609 Rev 4, Geneva, Switzerland, 2023.
**J. Zhang et al., “Research on unsupervised condition monitoring method of pump-type machinery in nuclear power plant”, Nucl. Eng. Technol., vol. 56, no. 6, pp. 2220–2238, 2024.

Funding Agency

Work supported by the U. S. Department of Energy, Office of Science, under Contract No. DE-AC02-06CH11357.

In which format do you inted to submit your paper? LaTeX
Preprint marking on your proceeding paper I do not wish my paper to be marked as preprint.

Author

Rajat Sainju (Argonne National Laboratory)

Co-authors

Joel Fuerst (Argonne National Laboratory) Michael Borland (Argonne National Laboratory) Osama Mohsen (Argonne National Laboratory) Robert Wright (Argonne National Laboratory) Yine Sun (Argonne National Laboratory)

Presentation materials

There are no materials yet.