Innovations in AEC

This entry is part 1 of 6 in the series April 2021

How autonomous data capture is changing the industry

I’m sure it feels the same in every industry – sometimes there are just too many acronyms to keep track of: AEC, LiDAR, VR, BIM, 3D, 4D, 7D – the list goes on and on for AEC professionals. I’d like to add ADC to the list – autonomous data capture. It’s a development that I think is particularly exciting for AEC professionals everywhere.

What is ADC?

Autonomous data capture is the use of autonomous devices to conduct 3D scanning for use in point cloud CAD software programs. It’s an evolution of the existing trend of using devices such as drones to aid in 3D scanning of sites – but while those devices are all currently driven by a human pilot, autonomous devices drive themselves. The benefits of this are numerous:

Safety: Autonomous vehicles can enter hazardous terrain without endangering the safety of any personnel on site. They can also make it easier to keep a construction site COVID-secure by removing the need for on-site scanning teams.

Reduced manpower cost: After the initial outlay of buying the systems, AEC firms using ADC will find they spend less on contractors visiting sites to conduct scans, making operations leaner overall.

Improved efficiency: Because ADC is autonomous, it means site capture can take place overnight, when workers are off-site. In this way, ADC makes reality capture workflows – and therefore the entire construction process – more efficient.

More frequent scanning: Since ADC doesn’t need people to operate it, it’s cost-effective to conduct scans more frequently than before. This enables teams to more closely track progress on sites, catch errors faster and provide fuller information to stakeholders.

How does it work?

ADC essentially works the same as non-autonomous 3D scanning, in that 3D scans are conducted of a site, point clouds are generated, which in turn can be used by point cloud to BIM software.

The difference, as we’ve said, is that those scans are conducted by autonomous devices. The most prominent example is Boston Dymanics’ Spot Robot, which has announced collaborations with Trimble, Hilti and Holobuilder, but there are numerous other examples including a tank-esque robot from Doxel and a similar robot from Scaled Robotics.

The different systems work in slightly different ways. For instance, Boston Dynamics has designed its dog-like Spot Robots to carry lidar scanners from a range of different providers, while Scaled Robotics has used the Autodesk Forge API to build both hardware and software together. All the systems, however, share the approach of programming the robot with a pre-defined path to follow around a site, conducting scans either at pre-defined waypoints or continuously.

Where has ADC been used?

It’s still early days for ADC, with most providers still running small-scale pilot projects. However, my reading has thrown up these examples, and the benefits ADC has brought to each project:

Denver International Airport: Using the Boston Dymanics Spot Robot with a Trimble X7 scanner, ADC was used during the pre-planning phase to capture as-built conditions, sent to point cloud CAD software on-site in real time. This process helped verify that the entire space has been captured, reducing the risk of return visits – an especially important capability in this type of projects where  access permits and logistics can cause delays. The ADC information was then compared with the BIM model of the design to ensure construction took place with no delays.

San Francisco International Airport: This was another deployment of the Spot Robot, this time with Holobuilder RC technology. Reports of the pilot indicated that the on-site team rebuilding Harvey Milk Terminal 1 could get the ADC system to work with only minimal training, generating accurate scans and saving time for the team. Contractor Hansel Phelps estimates it saved 4,680 hours annually using ADC rather than traditional photo capture methods.

Medical Offices in San Diego: This project from Doxel saw their ADC robot capture scans of the entire site, Kaiser Permanente’s Viewridge Medical Office, every day, each scan taking 4.5 hours. The project saw increased labor productivity of 38 percent, and the project as a whole came in in 11 percent under budget.

The key to reality capture’s evolution?

Even though reality-capture as a discipline is still relatively new to the AEC industry, ADC technologies represent an exciting evolution in the field. With autonomous vehicles able to scan construction sites much more regularly and efficiently than humans, reality capture data can now track the progress of a construction site far more accurately – helping to maximize the savings AEC firms can attain in labor costs and efficiency gains.

Of course, there are challenges that ADC will bring, which also need to be solved. Chief among those is how to handle the explosion of data that increased scanning provides. Especially on large sites, point clouds can each be hundreds of gigabytes large – making them hard to open and manipulate without high-powered (and expensive) computers.

With new scans sometimes being produced every day, a way to handle that information and present it to project teams that doesn’t involve heavy investment in high-spec computers will be essential.

One option for handling that is looking at alternatives to point clouds altogether – looking at ways to take that data and convert it into smaller file formats that still give users the insight needed to successfully work on projects. It’s the option PointFuse chose to investigate, generating intelligent mesh models that, from the point cloud data, are comprised of discreet selectable surfaces that can be easily manipulated and classified by the project team.

It’s certainly not the only solution to the challenge – but whatever solution firms deploy to handle increasing data volumes, ADC will help shape the future of AEC and reality capture.

Series NavigationThe Early Days of 3D Scanning: Part 10 >>

Leave a Reply

Your email address will not be published. Required fields are marked *