Machine learning based sensing of particle shape and size using passive artificial cilia

Published in Journal of Fluids and Structures, 2026

Passive artificial cilia on a channel wall are deflected by the flow and by particles passing over them. This paper uses immersed boundary simulations of cilia–particle interaction and an LSTM neural network to infer a particle’s size and shape from the cilia deflections, without optical measurement.

Read the paper (DOI: 10.1016/j.jfluidstructs.2025.104446)

Related: cilia–particle simulation framework · ECCOMAS 2024 talk

Recommended citation: Divyaprakash and A. Bhattacharya (2026). "Machine learning based sensing of particle shape and size using passive artificial cilia." Journal of Fluids and Structures, 140, 104446.
Download Paper

Leave a Comment