The benefits of error-detection technology

8 Dec 22

Investment in technology to facilitate early detection of errors can reduce costly rework and minimise impact on project schedules and quality, Stephen Henley from Mace told attendees at GIRI’s members’ meeting on 24 November.

Stephen explained how Mace is using augmented reality, drones, and even trialling robotic scanners on its projects to map build progress in real time, detect issues, and increase quality, cost and programme certainty. “The more you use these technologies, the better you can work out how to integrate them into your quality workflows,” he said.

Commercial-grade drones track progress and monitor quality control of facades on taller buildings through the use of high-definition images and videos. “We also use a 360° Open Space camera on the same route every day to capture a set of 360° images, creating a permanent record of installation at any point in time.”

Mace is using an Atom AR headset from XYZ Reality, integrated into its quality workflows. This is a safety helmet with a visor and head-up display that enables the wearer to take the entire project BIM model onto site. “You can see a virtual model in the field in real time. You can check where you are with your schedule sequence, use it in void mode to see through ceilings or floors to the services or cabling, or see the whole building being built in front of you as you’re moving through it.”

The benefits of using AR in this way include the ability to pick up errors, such as pile caps or steel members in the wrong places, before they cause major problems. “Because we are picking these things up before they go into the ground, or even before they go to site, we can resolve them more easily, reducing rework, disruption, and impact on the schedule,” Stephen explained.

He highlighted some examples of AR’s error-detection on a major project on a 50-acre site which involves simultaneous construction of two large buildings. AR is used in site inspections, surveying an area to pick up various elements such as steel alignment or services in the ceiling. “We’ve had 328 issues detected on inspections. In each case, these may or may not cause problems. In some instances, we use the helmet with a controller to record these as the permanent installation for the record model, rather than, for example, take the steel down. Other times, we’ve had to take corrective actions.”

In one example, a section of steel frame was found to be 100mm out of alignment, in another a length of pipe was a metre astray. Both could have caused major issues if they had not been picked up until the next trade came in, but as they were discovered at the point of installation, non-conformance notices could be issued and the errors resolved without major impact on the schedule.

A knock-on advantage of this early detection is that it creates a culture of greater diligence, said Stephen. Another is an improvement in information management.

“The system can work well in any environment, as long as your models are accurate, and have the right level of detail, and you have strict control about how they are updated. Make sure it is built into your quality and digital strategy and that people understand how and why it is being used so you have a proper workflow.”

It is a considerable investment, however. So how does Mace quantify its value? “We find a significant number of issues that otherwise wouldn’t be recorded or would become problems further down the line when we would have to put them right. We quantify every non-conformance against the issue by calculating what it would have cost to fix if it had been discovered later. There is major value here, but that is not the only benefit. It’s all about the greater cost certainty, better schedule certainty, and better quality certainty.”

Finally, Stephen talked about a data centre project in Ireland which was trialling a Boston Dynamics ‘Spot the Dog’ robot equipped with a Trimble X7 scanner. This recently achieved its first autonomous 10-hour scan of the data hall. “It goes to a certain point, scans the room, then moves to the next point and scans. The battery runs down after 1.5 hours, and it goes back to the docking station, puts itself back into the dock to charge, then goes back out on its own again. We get the full 3D imagery of the site, which is available through our project control centre.”

Interesting in attending our members' meetings? Become a GIRI member.

Sign up to our newsletter for the latest GIRI news and events.

 

Back Back