Point Clouds - What’s the Point?
For those unfamiliar, a point cloud is a three-dimensional set of fixed location points in space that reference the edge and surface points of objects in that space. These point sets, known as “point clouds,” are typically generated by using a 3D scanner of some kind.
The initial use of point clouds and scanners on projects was limited by the cost of the equipment and the shortage of people trained in its use. Like every technology, its cost has dropped over time, making this service, once reserved for very large-scale projects, available to virtually everyone. In addition to lower cost, new tools provide more spatial information and gather it in a shorter period of time than their predecessors, making point cloud scanning less of an event and more of an everyday construction site practice.
Environment Capture
Two leading technologies can be used for generating 3D models of an environment.
Photogrammetry starts with a photograph, video, or series of photographs and generates a model of the physical environment using software analysis of objects in the captured images. This process is used to generate Google Maps from aerial photos among other resources.
Light Detection and Ranging (LIDAR) uses a pulsed laser and software to generate a non-photorealistic 3D map of the environment. LIDAR is used for piloting many self-driving vehicles, and combined with photogrammetry, generates the explorable Google Maps street views.
Though both methods have the potential for use in design and construction, LIDAR-generated point clouds are now capable of capturing RGB data, and typically result in more reliable information about the space being measured and the objects in it at any point in time. As a result, LIDAR scans are more applicable for use in most construction projects.
The speed of LIDAR scanning to generate a point cloud of a space is impressive. Space area that can be captured accurately is set by the equipment being used, but the process of generating the capture can be as little as two to three minutes.
Use with BIM and Project Execution
Once generated, the cloud can be aligned with the design or construction models. Doing so allows for check of construction progress, confirmation that work is being conducted consistent with what was designed and specified, and even remote inspection of the quality of some forms of work. Doing so on a regular basis can also surface problems before corrections become more costly in terms of money or schedule. This bottom-line reassurance is meaningful to project managers and owners.
At a recent conference, I spoke with Kyle Barker, AIA, about this type of use during delivery. He performs point cloud scans of sites as a regular part of project management and quality assurance/control. The cost savings, early correction of mistakes, and peace of mind that the technology provides are well worth the minor investment in equipment and becoming familiar with the tools and what they deliver. As he put it, “There’s really no going back to a tape measure and photo-based site inspection once you've done it with a laser scanner.”
In addition to use during project execution, point clouds provide a fast and easy way to ensure that measurements, whether provided in a bidding package or taken by a contractor preparing a bid, are generated quickly, cheaply, and accurately. Accuracy of the dimensions of a space, the size of objects in the space that may need selective demolition, and the condition of existing work that will need to be renovated or replaced are all made more reliable and less prone to observation errors.
Project Closeout
Looking forward, it is easy to see how point cloud surveys taken by the contractor could be submitted to building code officials along with video and photographs as a substitute for physical inspection of many projects. This would enable greater time efficiency for code inspectors and provide a better record of code compliance for owners and other project participants.
For owners, a final scan of project spaces would provide the basis for very accurate and inexpensive “as-built” models. Once aligned with the design and construction models they could be the basis of one set of information available for facility maintenance and space planning.