Artificial intelligence set to shape utility’s future

Posted 17 July 2018

SewerOne leading Australian utility has been sending a scout into its sewers to collect data, but it can do a lot more than that – the Sewer Scout has been developed to remember locations and identify defects, promising a new age of utility-managed artificial intelligence.
 
Presenting at the Australian Water Association NSW Conference next month, Sydney Water Customer Delivery Officer Steve Barclay said the utility’s new technology was developed in a bid to reduce the risks involved in the Avoid Fail Sewer traverse inspections.
 
“Traversing entails sending people down into our large sewers and walk through for a condition assessment. Basically, we are looking for three things: issues we need to fix now, soon or later, so that we can budget and prepare for works,” he said.
 
“Sydney Water has 16 fatal-risk standards and traversing includes 13 of those risks. Traversing is the riskiest operation undertaken at Sydney Water. Our systems are quite robust, but still it is a high-risk activity.”
 
Barclay said Sydney Water has been looking for a technology-based solution for this issue, with recent developments in geolocation and other information technology offering a pathway.
 
“We’ve been looking for a system that can gather enough information for detailed engineering assessments without having to traverse. And technology is finally catching up with us,” he said.
 
“Sewer Scout is the result of this technology crossover. It uses the latest in photogrammetry  – I like to call it Google Street View for sewers – stitching lots of photos together to come up with a 3D model. You can view the conduits as if you were there yourself, but without having to get dirty.
 
“The only thing you can’t do is take tactile measurements, but we can undertake these tasks at the launch or retrieval maintenance holes. However, it reduces our risks right down to three of our 16 fatal risk standards.”
 
Furthermore, as the Sewer Scout retrieves complex data, it learns how to locate and identify defects, Barclay said, heralding in a new era of artificial intelligence for Sydney Water’s operational and maintenance decision making.
 
“The Sewer Scout self-centres in the sewer flow – it collects the images with a special floating camera arrangement which are then converted into 3D models and a fly-through view,” he said.
 
“The images are captured and downloaded into the software, which is now being developed to take the Scout to the level of artificial intelligence. The Scout will be able to locate and identify defects in the system. We are at the stage where it is 98% accurate in locating maintenance chambers and holes.”
 
Barclay said the sophisticated technology will play a role in the utility’s future asset management, but for now Sydney Water is pleased with how well the Sewer Scout helps in improving safety and reducing cost.

“This technology will help us prioritise our future works. We assess the service life of every Avoid Fail Sewer asset we have on the traverse program. We want to try and automate the entire system,” he said.
 
“We will always have to go down into sewers to work on them and check things we are unsure about. But this will eliminate the majority of our current risk profile.”
 
Register for the Australian Water Association NSW Conference to hear more from Steve Barclay about Sydney Water’s Sewer Scout technology.

Related articles:
AI could be the future of decision-making in the water industry