Concrete infrastructure including bridge structures in coastal areas can experience faster deterioration posing a threat to the users of the infrastructure. In order to ensure safety, infrastructures including concrete bridges need to be periodically inspected for defects and more importantly their changes from previous inspections. This process is labour intensive and normally requires the closure of the infrastructure during these inspections. Therefore, digital automated techniques are desirable to reduce inspection times, avoid infrastructure closures, and reduce safety risks to inspection personnel. Wall climbing robots provide a possible solution to automate the capture of imagery of the concrete structure. This data is then processed and analysed using computer vision and machine learning techniques to understand if there are new defects or if there are any changes to previously identified defects, such as cracks. The output is then mapped to a virtual reality environment through which surveyors can inspect the health of the bridge and evaluate risks.

Coastal bridges play a crucial role in transportation infrastructure, connecting regions and facilitating the movement of people and goods. However, due to their exposure to harsh coastal environments, these bridges are susceptible to various forms of deterioration, such as corrosion and crack. Traditionally, the inspection and evaluation of coastal infrastructure including bridges have been highly challenging and time-consuming tasks, often requiring human inspectors to access hard-to-reach areas at great risk. Therefore, there is an urgent need to develop innovative and effective digital methods for detecting and assessing the damage of coastal bridges. The primary objective of this project is to utilize climbing robots and computer vision technologies for the inspection and assessment of coastal bridges. The project's specific objectives are as follows:

1. Develop a novel approach for movement trajectory planning of climbing robot for coastal bridges by integrating Building Information Modelling (BIM) and computer vision techniques. The trajectory planning should ensure efficient and safe data collection of the bridge damage assessment.
2. Develop a deep learning-based method for recognizing and quantifying the bridge damage using a fusion of image and point cloud models. Propose rapid damage recognition and quantification
   by integrating 2D images and 3D point clouds to improve damage detection accuracy and realize damage localization and interactively visualization. Develop automated structural assessment based on Industry Foundation Classes (IFC)-based semantic modelling and DT model.
3. Automated modelling based on 3D point clouds. The objectives include investigating automated and accurate methods for point clouds segmentation; automated fitting and modelling algorithms to generate as-is 3D models based on the segmented point clouds.
4. Automated establishing and updating BIM system of coastal bridges for structural assessment. The objectives include developing a Finite Element (FE) model incorporating the detected damages for predicting the residual life of existing structures; a BIM system of coaster bridges, which helps to predict the structural performance and provides infrastructure operators with a decision-making support tool for an efficient management of the structure under normal and extreme loading conditions.
   
   The proposed project aims to develop a novel approach for detecting and assessing the damage of coastal infrastructure, including bridges, using a combination of climbing robots and computer vision technologies. The project's objectives are realistic and achievable, and the outcomes will have a significant impact on the field of coastal bridge health monitoring. The success of the project will be evaluated based on the accuracy of the damage detection and assessment of the in-service coastal bridge.