Localization of Impacts on Thin-Walled Structures by Recurrent Neural Networks: End-to-end Learning from Real-World Data

2025-05-13

AI Summary

Accurate impact localization on thin-walled shell structures remains challenging due to complex wave propagation and signal dispersion. To address this, we propose an end-to-end Gated Recurrent Unit (GRU) model trained exclusively on high-fidelity, robotically generated impact data from piezoelectric sensors, directly regressing 2D impact coordinates from raw high-sampling-rate time-series signals. Our key contributions are threefold: (1) the first use of fully automated, physically realistic impact data—eliminating domain shift inherent in synthetic waveform generation; (2) a lightweight GRU architecture specifically designed for long temporal sequences (thousands of timesteps), enabling efficient processing of fine-grained sensor signals; and (3) millimeter-level localization accuracy achieved with only a small-scale real-world dataset—substantially outperforming conventional Lamb wave-based time-of-flight and waveform analysis methods. This work demonstrates the feasibility and superiority of end-to-end deep learning for impact localization in structural health monitoring.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Machine Learning: Multimodal Learning
• Planning, Routing, and Scheduling: Activity and Plan Recognition

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Today, machine learning is ubiquitous, and structural health monitoring (SHM) is no exception. Specifically, we address the problem of impact localization on shell-like structures, where knowledge of impact locations aids in assessing structural integrity. Impacts on thin-walled structures excite Lamb waves, which can be measured with piezoelectric sensors. Their dispersive characteristics make it difficult to detect and localize impacts by conventional methods. In the present contribution, we explore the localization of impacts using neural networks. In particular, we propose to use {recurrent neural networks} (RNNs) to estimate impact positions end-to-end, i.e., directly from {sequential sensor data}. We deal with comparatively long sequences of thousands of samples, since high sampling rate are needed to accurately capture elastic waves. For this reason, the proposed approach builds upon Gated Recurrent Units (GRUs), which are less prone to vanishing gradients as compared to conventional RNNs. Quality and quantity of data are crucial when training neural networks. Often, synthetic data is used, which inevitably introduces a reality gap. Here, by contrast, we train our networks using {physical data from experiments}, which requires automation to handle the large number of experiments needed. For this purpose, a {robot is used to drop steel balls} onto an {aluminum plate} equipped with {piezoceramic sensors}. Our results show remarkable accuracy in estimating impact positions, even with a comparatively small dataset.

Problem

• Localize impacts on thin-walled structures using RNNs
• Overcome Lamb wave dispersion challenges in impact detection
• Train networks with real-world data via automated experiments

Innovation

• Uses recurrent neural networks for impact localization
• Trains with physical data from automated experiments
• Employs Gated Recurrent Units to handle long sequences