Sign in Subscribe
Periodical publishing

Artificial Intelligence Improves Structural Health Monitoring in Large-Scale Structures

Strategic Insights

— 1 min read

Researchers at Inha University have developed an artificial intelligence (AI) system that uses machine learning algorithms to monitor the structural health of large-scale structures such as bridges and buildings. The system, which includes a flexible temperature sensor fabricated using EHD inkjet printing, was tested on dynamic data and achieved a 33.563% improvement in prediction error over conventional polynomial regression. The AI system demonstrated superior generalization, reducing the Root Mean Square Error (RMSE) from 12.451 °C for the polynomial model to 4.899 °C.

Key Takeaways:

  • The research team at Inha University developed a flexible temperature sensor using EHD inkjet printing to monitor thermal effects in large-scale structures.
  • The AI system, which uses a Long Short-Term Memory (LSTM) calibration model, achieved a 33.563% improvement in prediction error over conventional polynomial regression.
  • The LSTM model was trained exclusively on quasi-static data across the 20-70 °C temperature range and demonstrated superior generalization when tested on unseen dynamic data.
  • The AI system reduced the RMSE from 12.451 °C for the polynomial model to 4.899 °C, suggesting that data-driven approaches like LSTM can be highly effective solutions for ensuring the reliability of flexible sensors in real-world SHM applications.
  • The research was funded by the Inha University Research Grant and published in the journal Sensors.
  • Additional authors for the research include Ju-Hun Ahn, Ji-Han Lee, and Chang-Yull Lee.

Statistics:

  • The AI system achieved a 33.563% improvement in prediction error over conventional polynomial regression.
  • The LSTM model was trained on quasi-static data across the 20-70 °C temperature range.
  • The AI system reduced the RMSE from 12.451 °C for the polynomial model to 4.899 °C.
  • The research was published in the journal Sensors on October 13, 2025.

Sources:

  • "Temperature Calibration Using Machine Learning Algorithms for Flexible Temperature Sensors." Sensors, vol. 25, no. 18, 2025, pp. 5932. (Sensors - http://www.mdpi.com/journal/sensors)
  • NewsRx. "Inha University Researchers Highlight Research in Machine Learning (Temperature Calibration Using Machine Learning Algorithms for Flexible Temperature Sensors)." Journal of Engineering, October 13, 2025, p 1189.

Read more