Smart bridge health monitoring and maintenance prediction
This project aims to investigate the feasibility of using an integrated package of IoT, computer vision, and machine learning technologies to support smart bridge health monitoring and prediction.
Integrated IoT, computer vision, and machine learning technologies offer a promising supplement to physical bridge health assessment particularly in remote regional contexts which can be costly, time consuming and unsafe to inspect. Conducting regular, efficient, and reliable bridge health monitoring is essential for the long-term protection of valuable road assets through timely maintenance responses.
The research from this project will produce a proof-of-concept to demonstrate the efficacy and feasibility of an integrated package of technologies for first-level bridge health screening and early warning system, reducing the need for traditional physical inspections and instrumentation.
The benefits of the project include contributing to reducing maintenance, operation costs and risk, and achieving a safe transport infrastructure network, ultimately, increasing productivity.
In Western Australia there are close to 3,000 bridges with Main Roads WA owning the largest proportion of bridge structures (42%), thus responsible for the routine inspection and maintenance of these assets. Further, Main Roads WA is responsible for maintaining truck loading standards for publicly accessible roads and bridges, owned primarily by Main Roads WA or local authorities.
The vast majority of bridge structures are located at considerable distances away from major urban centres making regular physical inspections difficult and costly. This could mean missed windows of opportunity for optimal treatment when structural problems occur, which could lead to higher overall maintenance costs and increased risks of major structural failures. It could also mean more disruptions to freight movement during maintenance.
Reinforced concrete and prestressed concrete bridges, which form a large and growing proportion of the bridge stock (51%) in WA, have their own unique set of instrumentation and health evaluation challenges. This research will specially consider concrete bridges where the opportunity for solutions which reduce physical instrumentation requirements, further justifies the importance of the proposed research.
Bridge displacement, vibration and traffic load are three key indicators in bridge health monitoring. Level of displacement is a good descriptor of structural deformation behaviour and an indicator of structural performance and has been used widely for bridge health monitoring. Owing to high instrumentation cost for conducting bridge inspections and vibration tests, computer vision-based methods provide a cost effective and efficient alternative for measuring bridge displacement responses and conducting bridge condition monitoring.
IoT technologies can aid in timely data collection resulting in real time bridge health monitoring and facilitating timely generation of warning alarms. The computer vision, IoT, and AI-based smart decision-making tool will enable Main Roads WA to monitor bridge health conditions on an ongoing basis, which not only saves on inspection costs but also increases the efficiency of the maintenance program. It will generate data that has been previously unavailable for better modelling how bridges respond to traffic loading.
The results could inform whether current truck loading guidelines need revision to achieve the best balance between maintaining bridge health and maximise the long-term efficiency of the freight industry.
The data could also help increase compliance and avoid overloading within the freight industry, which ultimately benefits all stakeholders.
The overall aim is to demonstrate the feasibility of using a combination of IoT, computer vision, and machine learning technologies to reduce the need for manual bridge inspections and provide on-going monitoring that supports smart bridge maintenance.
In this project, the development, application, and integration of IoT technology-based sensor data collection, computer vision methods and machine learning techniques will be investigated to enable remote monitoring of bridge structural health to enhance the efficiency of asset monitoring, management, and maintenance.
The specific project objectives are to:
- Develop IoT technology-based solutions for collecting, visualising, and transmitting vibration response data.
- Develop computer vision-based solutions for measuring bridge displacement under a range of traffic loads.
- Develop an integrated proof-of-concept predictive model based on machine learning techniques for predicting bridge displacement under different traffic loads; and
- Make recommendations on new techniques for efficient bridge monitoring, wider roll-out of the technology for bridge health monitoring and prediction and suggest the next steps.
Please note …
This page will be a living record of this project. As it matures, hits milestones, etc., we’ll continue to add information, links, images, interviews and more. Watch this space!