Automated asset inspection unlocks miles of data and speeds up pipeline repair

We’re using artificial intelligence to help speed up repairs to our network of wastewater pipes by semi-automating the review of CCTV inspection footage.

Since the use of CCTV for pipeline inspections was first introduced in the 1950s, the process of reviewing footage has remained largely manual – and very time consuming. The increasing volume of work across the water sector has highlighted the need for modernisation. Every year we inspect more than 1,000km of pipeline using CCTV cameras – and that generates thousands of hours of footage which until now has been manually reviewed to check for pipe damage or obstructions which can lead to flooding. Now we have adopted cutting-edge technology to dramatically speed up the process.

Thanks to a successful partnership with VAPAR, an Australian artificial intelligence (AI) company, CCTV survey processing times have been reduced from 10 days to two days – an 80% improvement.

VAPAR uses AI to speed up pipeline repairs by semi-automating the fault detection process from inspection footage. Founded by two female engineers in Australia, we first began working with VAPAR in the second Innovation Lab held in 2019. Initially the focus was on developing the accuracy of the AI.

The complex nature of pipeline inspections makes it impossible for AI to be used as a stand-alone tool and the outputs still require review. VAPAR uses a ‘Collaborative Intelligence’ approach which sees AI being used to support human expertise.

To find out more, head to our news section.