Use of Artificial Intelligence for Analyzing Structural Health Monitoring Data

Title: Use of Artificial Intelligence for Analyzing Structural Health Monitoring Data from Concrete Structures
Presented By: Harshita Garg, University of Leeds
Description: It is now straightforward to collect data concerning service conditions and performance of building materials in structures. However, accurate and efficient approaches to interpret large volumes of uncertain data from such real-time monitoring systems are [not well-established in the field of structure health monitoring (SHM). Consequently, timely diagnosis using the SHM data is still a challenge and, hence, making a proper decision for structural repair and maintenance is currently a challenge using the raw data. This study proposes an Artificial Intelligence (AI)-enhanced methodology to interpret the reliability of SHM data and obtain the standard steady-state condition to predict the performance of three different types of concrete. A large amount of data was collected continuously from high-performance concrete blocks embedded with a multielectrode array for acquiring electrical resistance and thermistors for temperature measurements. An AI-algorithm was developed to reduce the influence of noisy data and perform ‘segmented regression’ for an effective and reliable interpretation of the SHM data. The algorithm divides the points into segments such that the slope difference was the greatest, thereby providing the intersection points for the segments to determine the ultimate breakpoint for change in resistance, or the diffusion coefficient. The results demonstrated that the proposed methodology is faster and efficient than the conventional approach to calculate the stabilised value of resistance/diffusion coefficient. Data from the sensor further highlighted the impact of temperature and depth on resistance measurements for different concrete mixes. In addition, it showed that the use of supplementary cementitious materials in concrete has long-term benefits for providing the resistance to the ingress of aggressive substances into the concrete.