AI to prevent bridge collapses

Five people were injured in 2017 when a bridge collapsed onto a Johannesburg highway. Fortunately, one company is preventing bridges from collapsing with the help of 3D imagery

Dynamic Infrastructure is implementing the first deep-learning solution, which assists owners and operators to obtain a visual diagnosis of bridges and tunnels. The system provides live, cloud-based, 3D views with automatic alerts when changes are detected in maintenance and operational conditions before the issues evolve into large-scale failures.

The start-up is already running projects in the Unites States (US), Germany, Switzerland, Greece and Israel with different transport infrastructure stakeholders.

The system quickly creates records for bridges, tunnels and elevated highways, based on existing images that have been taken through periodic condition inspections over the years, including images from smartphones, drones and laser scanning. The proprietary technology compares old and archived images to new ones in order to detect maintenance and operational issues, defects and anomalies.

Similar to the magnetic resonance imaging used for diagnosing health issues in people, the 3D records serve as the basis for the alerts on changes in maintenance conditions. The diagnostics can be easily accessed through a simple browser and shared with peers or contractors to speed up maintenance.

“The world faces an infrastructure crisis,” says Saar Dickman, co-founder of Dynamic Infrastructure. “Specifically, deficient bridges and tunnels represent a severe infrastructure challenge in the US and worldwide. Their poor condition can lead to loss of life and can cost millions in terms of unplanned expenditures.”

He adds that attempting to repair infrastructure without implementing new technologies won’t resolve the issue. Through the use of technology and preventative maintenance, companies can reduce expenditure and potential risk.

“A single dollar spent on the right technology in the right place can save much more than one dollar spent on the maintenance of a bridge,” he explains. “Until recently, there has been no effective system that can quickly and precisely identify defects in bridges throughout their lifetime.

“Dynamic Infrastructure provides actionable monitoring and alerts that can help to better manage expenditures and prevent the next collapse,” Dickman continues. “Our cutting-edge imagery analysis is bringing the data revolution to the decision-making process of maintenance of bridges and tunnels.”

In recent years, South Africa has been plagued by the collapse of several pedestrian bridges. In 2017, five people were injured when a bridge collapsed along the N3 Geldenhuys Interchange in Germiston. Even more devastation occurred in 2015 when a bridge collapsed over the M1 highway in Sandton. Two people died and 19 were injured.

Through imaging technology, companies can predict maintenance and protect its asset but also prevent the loss of life that can occur when a bridge caves in.