International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056 p-ISSN: 2395-0072
AI-Driven Predictive Maintenance of Structural Systems via MobileNetV2-Based Damage Classification Rikesh Srivastava1, Abhishek Mishra2, Sachin kumar singh3 r1Student , Dept Of Civil Engineering ,Iet Lucknow India 2 Assistant Professor, Dept Of Civil Engineering ,Iet Lucknow India
3 Assistant Professor, Dept Of Civil Engineering ,Iet Lucknow India ---------------------------------------------------------------------***--------------------------------------------------------------------intensive, and unable to scale effectively for large Abstract - The integrity and serviceability of structural structures or networks. With the growing interest in systems are increasingly threatened by aging, smart infrastructure, there is a compelling need for environmental exposure, and insufficient maintenance predictive and automated systems that can assess practices. Conventional inspection techniques, while structural health with minimal human intervention. widely used, are labor-intensive, time-consuming, and often reactive rather than proactive. This study proposes a predictive maintenance framework using MobileNetV2 Recent advancements in Artificial Intelligence (AI), (MobileNetV2) to automate the detection and especially in Deep Learning, have revolutionized many forecasting of structural surface defects. A customengineering domains. MobileNetV2 (MobileNetV2), a labeled image dataset comprising various damage class of deep learning models particularly effective in types— Intact Crack Spalling Corrosion, is utilized to image processing tasks, offer a promising pathway train the MobileNetV2 model. The model not only toward real-time and scalable Structural Health achieves high accuracy in multi-class damage Monitoring (SHM). By leveraging historical image data classification but also demonstrates potential in and learning visual patterns associated with recognizing early patterns of deterioration through degradation, MobileNetV2 enable early detection of sequential image analysis. Performance evaluation defects and classification of their severity levels. Unlike indicates superior detection precision compared to traditional models that focus purely on object traditional deep learning models such as Faster Rdetection, MobileNetV2 excel at classification tasks MobileNetV2 and YOLOv5, with reduced computational where nuanced differences in visual features (such as requirements. The findings suggest that the proposed crack depth or spalling extent) are critical. MobileNetV2- based approach can significantly enhance structural health monitoring by facilitating timely In this research, we present a MobileNetV2-based interventions and extending the lifespan of predictive maintenance system designed to identify infrastructure. This study contributes toward the and classify surface defects in civil structures. Our advancement of AI-driven predictive maintenance approach utilizes a custom image dataset, carefully systems and their application in real-world civil annotated and restructured from YOLOv5 format into a engineering contexts. MobileNetV2-compatible structure. The images Key Words: Predictive Maintenance, MobileNetV2, Structural Health Monitoring, Deep Learning, Damage Classification, Infrastructure, Civil Engineering
1.
INTRODUCTION
The increasing complexity and aging of global infrastructure necessitate innovative approaches to ensure safety, longevity, and cost-effective maintenance. Conventional inspection and maintenance practices, such as manual visual inspections or non-destructive testing methods, are inherently reactive, prone to human error, laborImpact Factor value: 8.315
represent a spectrum of damage severity: intact, crack, spalling, and corrosion. Our results demonstrate that the proposed model not only achieves high classification accuracy but also holds the potential to identify deterioration trends when extended with sequential learning methods. This study emphasizes the potential of AI-driven systems in transforming infrastructure maintenance strategies from reactive to predictive, enabling civil engineers and city planners to make informed decisions for extending the service life of critical assets
ISO 9001:2008 Certified Journal