Saturday 26th of May 2018
workflow4point0 Logo

Tips, tactics, ideas and opinions to optimize your 3D modeling workflow

Subscribe to Workflow 4.0

Share this Newsletter

Cannon Saves 60% with Automated Feature Extraction on an Urban Rehab BIM
Written by Kevin Corbley   

The former West Coast headquarters for Coca-Cola in Los Angeles is undergoing a major conversion. Once a cavernous soft drink production facility, the warehouse-style building will soon be home to offices, retail stores and restaurants. But the make-over hasn’t taken place without the structure revealing a surprise that posed a challenge to 3D BIM creation. “The cement floors in the building were not flat – there were raised areas and then slopes down, almost like it was purposely tiered to provide level platforms for some type of processing equipment or storage tanks” said

Lester Carter, Director of Surveys at Cannon Corp., a full-service engineering, GIS and 3D modeling firm based in San Luis Obispo, Calif.


Carter could only speculate that the floor was designed that way for a reason. Perhaps the production equipment was placed on the higher levels so that spills would drain away quickly. Or maybe the sloping floor and apparent “pathways” made it easier to cart away loads of bottled product. Whatever the reason, the cement floor had its own topography that had to be modeled. “The cement sloped as much as two inches over 12 feet [horizontally] in some areas,” said Carter. “The designers needed an accurate as-built model of the floor surface to accurate design the interior construction of the building.”

150 Scans Over Three Floors

The developer hired Cannon to scan the gutted interior and provide a 3D REVIT/BIM model containing architectural, structural and MEP features. Specifically, they wanted as-built details for floors, walls, ceilings, support columns and beams, stair wells and existing electrical conduits and water and fire sprinkler piping. Over a period of six days, the Cannon team used FARO Focus 3D X-330 scanners to capture 150 scans for the basement, first through third floors and roof level.

After processing and registering the point clouds, Cannon ran them through the EdgeWise software for automated extraction of features. Typically, the firm’s technicians ran separate extractions for architectural, structural and MEP features using the EdgeWise feature extraction routines. These each ran overnight, floor by floor. “We estimated the automated software saved us 40-60 percent on extraction of 3D features over the course of the project,” said Carter. The results were outstanding. Cannon had promised the design firm the 3D models of the interior floor surfaces of the building to an accuracy of within 3/8ths of an inch, and they were. Carter noted that with additional scans for a higher level of overlapping scans, and supplemental scan targets, they could have as built the concrete floors to an even higher degree of accuracy.

Impressive Accuracy and a 60% Workflow Savings

Another challenge were vertical cement columns that supported the weight of the floors above. These were located throughout the building on every level. While EdgeWise identified and modeled the columns themselves, manual editing in Revit was required to model the geometry of the unique cone shapes and ceiling support plates at the tops of the columns. “We compared these and other features to the full-color point cloud and 360° photos taken by the FARO x330, for QA/QC purposes.

The results were, impressive” said Carter. Cannon delivered a variety of final products in Revit to the design firm, most notably a 3D BIM Model containing all of the relevant features needed to design, engineer and entirely re-build the interior of the building. You can learn more about this case study on a webinar this Wednesday 7/22 at 1pm EDT titled: As-Built BIM and Urban Renewal: Case Studies in Mixed-Use Redevelopment. Click here to register.


Add comment

Security code

Copyright © 2018 All Rights Reserved.
Joomla! is Free Software released under the GNU/GPL License.
Design by Joomla Bamboo