Autonomous Drone Surveys and Convolutional Neural Networks for Bridge Maintenance: A Predictive Approach Using Finite Element Analysis

The primary objective of this research is to develop a comprehensive framework for bridge maintenance that integrates autonomous drone surveys, convolutional neural networks (CNNs), and Finite Element analysis.

This approach aims to identify structural weaknesses in bridges, particularly cracks, and use this information to update a Finite Element model of the bridge. The updated model can then be used to predict when intervention is required, shifting the paradigm from reactive to condition-based maintenance.

Background on bridge maintenance

Maintaining the structural integrity of bridges is a complex task that requires detailed and accurate data on the bridge's condition. Traditional methods of data collection, such as manual inspection, can be time-consuming, costly, and may not capture all the necessary details. Furthermore, many bridges are ideal candidates for creating 3D meshes from surface meshes as they typically do not have significant internal voids, unlike buildings.

Use of Autonomous Drones

Autonomous drones can help overcome many of the challenges associated with data collection for bridge maintenance. They can easily access hard-to-reach areas of the bridge, capturing high-resolution video footage from a variety of angles. This footage can then be analysed using convolutional neural networks to identify cracks and their severity.

Convolutional Neural Networks for Crack Identification

Convolutional neural networks (CNNs) are a type of deep learning algorithm that can analyse visual imagery. In this research, a CNN will be trained to identify cracks in the drone footage, providing detailed information about their location and severity. This data will then be used to update the finite element model of the bridge.

Finite Element Analysis for Predictive Maintenance

Finite Element analysis is a numerical method used for predicting how a structure will react to certain forces, vibration, heat, and other physical effects. In this research, a Finite Element model of the bridge will be updated based on the data from the CNN. This updated model can then be used to run analyses and predict when and where intervention is required, enabling a shift towards condition-based maintenance.

Potential Impact

This research has the potential to revolutionise the field of bridge maintenance. By integrating autonomous drone surveys, convolutional neural networks, and Finite Element analysis, it could significantly improve the accuracy and efficiency of bridge inspections, and enable more effective maintenance strategies. Furthermore, the methods developed in this research could have broader applications in the field of civil engineering, potentially transforming the way we maintain and repair all types of structures.

Ideal Candidate

The ideal candidate for this project is a dedicated and innovative thinker with a strong background in Structural Engineering or related fields. They must have proficiency in Finite Element Modelling, and an interest in IoT, drone operations and image processing will be an advantage. The candidate should have strong analytical skills and the ability to work independently. They should also have a keen interest in bridging the gap between traditional structural engineering practices and emerging technologies. Experience with or willingness to learn machine learning methods is also desirable.

Brunel offers a number of funding options to research students that help cover the cost of their tuition fees, contribute to living expenses or both. See more information here: https://www.brunel.ac.uk/research/Research-degrees/Research-degree-funding. The UK Government is also offering Doctoral Student Loans for eligible students, and there is some funding available through the Research Councils. Many of our international students benefit from funding provided by their governments or employers. Brunel alumni enjoy tuition fee discounts of 15%.