According to The Australian Financial Review, GHD, a major engineering and construction company in Australia, cut 80 per cent off the cost of inspecting a bridge for Queensland Rail using a drone.
Using the drone with its fixed camera, the GHD engineer was able to take 21,000 images of the ageing steel structure spanning a creek, permitting the single-span bridge to remain open, rather than having to close it for the five or six days needed for a full inspection by humans.
"When you combine that with artificial intelligence, you can teach the software what to look for," Bugeja says. "So when we're doing inspections, you can say to the software, 'This is what a crack looks like', or 'This is what corrosion looks like'. The software starts to learn what we are looking for."
The technology is even more valuable when images the drone-mounted camera takes and the notes the consultant makes – typically resulting in a report stretching for hundreds of pages – are attached to a 3D model of the bridge, and allow engineers and their clients to study them from a desktop, or even in the field on a tablet.
"So rather than handing over a 540-page report and condition assessment, the client can just open up the model, they'll see a highlighted part of the model where there's an issue, they'll see a high-definition image of that state, and the comments associated with it," Bugeja says.
While drones and cameras can help in the construction of new buildings by photographing ground surfaces and measuring topography ahead of detailed design, they can arguably be more valuable in long-term asset maintenance. It's not just bridges, but inspecting the inside of exhaust stacks at power plants.
Ref:http://www.afr.com/business/construction/drones-using-ai-bring-productivity-gains-for-engineering-and-construction-20170511-gw2v40
If you want to read this article in Japanese, please see the following link:
https://www.j-abc.com/jp-blog/ai
Using the drone with its fixed camera, the GHD engineer was able to take 21,000 images of the ageing steel structure spanning a creek, permitting the single-span bridge to remain open, rather than having to close it for the five or six days needed for a full inspection by humans.
"When you combine that with artificial intelligence, you can teach the software what to look for," Bugeja says. "So when we're doing inspections, you can say to the software, 'This is what a crack looks like', or 'This is what corrosion looks like'. The software starts to learn what we are looking for."
The technology is even more valuable when images the drone-mounted camera takes and the notes the consultant makes – typically resulting in a report stretching for hundreds of pages – are attached to a 3D model of the bridge, and allow engineers and their clients to study them from a desktop, or even in the field on a tablet.
"So rather than handing over a 540-page report and condition assessment, the client can just open up the model, they'll see a highlighted part of the model where there's an issue, they'll see a high-definition image of that state, and the comments associated with it," Bugeja says.
While drones and cameras can help in the construction of new buildings by photographing ground surfaces and measuring topography ahead of detailed design, they can arguably be more valuable in long-term asset maintenance. It's not just bridges, but inspecting the inside of exhaust stacks at power plants.
Ref:http://www.afr.com/business/construction/drones-using-ai-bring-productivity-gains-for-engineering-and-construction-20170511-gw2v40
If you want to read this article in Japanese, please see the following link:
https://www.j-abc.com/jp-blog/ai
RSS Feed
