Ageing bridges monitored for impending damage using predictive AI tool

Korea Institute of Civil Engineering and Building Technology (KICT) researchers have developed a new technology to predict the levels of damage and ageing of bridges.

The DNA (Data, Network and AI) could be used for preventive maintenance and to thus avoid possible disasters. 

The scientists were aiming to address a significant Korean reality: the fact that the percentage of Korean bridges aged 30 years or more is expected to increase in the next decade to 39.3 per cent by 2031 and spike up to 76.1 per cent in 20 years.

With a view towards helping track damages, the KICT research team garnered more than five million data elements either directly or indirectly related to the ageing of bridges from 2021 to 2022.

The researchers then applied an artificial intelligence (AI) model to the data, to predict the degree of obsolescence of the bridges and predict damage over time.  

The credibility of the technology was further improved by securing additional data on the bridges using IoT technology onsite and from experimental data that considers the environmental conditions of Korea.

The technology was to have a 90.8 per cent predicted accuracy as of the end of 2022, which is expected to be further improved up to 95 per cent during 2023.

Dr. Ki-Tae Park, lead researcher, commented: “Securing preventive maintenance information on bridges by using various data, AI and network-based platform technologies will contribute to the longevity of bridges”, adding that, “the prediction services will expand in the future to include not only bridges but a wide range of infrastructure.”

The team at KICT said they plan to make the predictions publicly available via a platform in the second half of 2023. 

The BMAPS (Bridge Maintenance-Aided Platform Service) platform will offer prediction results and diverse bridge maintenance information services, such as load-carrying capacity (the ability to support weight) predictions for ageing small and medium-sized bridges.

The platform will initially be offered in Korean, with an English version to be prepared by 2024 which could support international calls for a paradigm change” in the science of forecasting corrosion damage in important structures such as bridges.

Last year, E&T found that councils hit hardest by the floods in late 2019 and 2020 faced a combined bill of half a billion pounds to fix bridges in need of repair. After the floods, local authorities warned of “lasting effects” and called for additional funding from the national government.

In 2020, the RAC Foundation revealed that the UK has  3,000 substandard road bridges in use, which are not capable of supporting the heaviest vehicles.

Four years prior, Swedish construction engineers embraced IoT technologies to address this issue by fitting sensors into bridges, making the infrastructure not only self-monitoring but also able to tweet about its condition.

The KICT project was detailed in a paper published in the journal Applied Sciences.

Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.

Original Source:

Action restricted!