Scan to BIM: The Evolution of Scanning Technology

The simple truth is, scanning may be the only cost-effective solution to collect the existing world.

You simply can't get into a cathedral, petroleum refinery, or metropolitan multi-use entertainment facility and measure with rulers and be prepared to get the accuracy you need to confidently design renovations.

Laser scanning is the only way to do it.

Until recently, BIM users would have a group of "asbuilt drawings" put them right into a 3D modeling program and develop a computer model to work from. Now, after several years to do that, the harsh realization has surfaced that there are many discrepancies between the "record drawings" and the actual environment to be constructed.

Whether it's sheetrock and wood, it could be adjusted to fit. But whether it's glass, steel, concrete or mechanical equipment, a seemingly small error can grow very costly as it is much harder to warp and bend. (Putting expensive new equipment into an area that is too small is a nightmare for the installer, designer, engineer and the insurance company.)

These new 3D laser scanning technologies have dramatically changed the surveying industry - and they have changed it fast. But to essentially understand the evolution, let's take a step back....

2004: 360-Degree Scans

The initial 360-degree scanners came onto the scene around 2004. Before Extra resources|Find more info , in the event that you wanted to scan something above your head, you had to tilt the scanner back and scan at a steep angle, because so many only had a 120-degree scan ability on the vertical axis. Several companies came out with full straight scanners about this time that made it much easier.

2006: Time-of-Flight Scans

Another evolution was time-of flight scanners. In 2006, a time-of-flight scanner took about 45 minutes to one hour for a complete 360-degree scan. In the event that you could do 8-10 scans each day, you were doing perfectly. Today, the same can be carried out in about 12-15 minutes, based on the density you will want scan.

At our firm, our first scanning projects were roads. In a very complicated area, we would scan 1"X 1". The time-of-fight scanners back then could collect 4,000 points per second. Now they are able to easily collect 50,000 points per second!

2008: Phased-Based Scans

Today's phase-based scanners collect 2,000,000 points per second and can develop a �-inch x �-inch pattern at a distance of about 100 feet. That is incredible so when fast and dense because the average user needs. The hardware will eventually progress, faster and cheaper, but phase-based scanning works well, stable, and provides the ability to scan just about anything in an acceptable about of time.

Present: Scan to BIM

Today, the big research money is going towards Scan to BIM technology, which converts vast amounts of points in the idea cloud into useful data.

Several companies have begun addressing this including small independent companies like Pointools, which came up with a means for scanners to identify flat surfaces. (As small as this may seem, it is just a huge advancement.) The program may also recognize pipes and model them automatically about 50% of the time. (Another major advancement.)

Now lots of the pipe programs are getting to the same place and advancing the ball. Currently, we have been at what I call the "Model T Ford" in software packages, but every year the programs progress.

The next evolution

Having now scanned may highly complex areas in industrial sites, we have had a chance to compare them to the asbuilt drawings. In the horizontal view, they are generally close geometrically to the actual. However in their vertical axis, the pipes and duct work in the asbuilt drawings are rarely correct.

There are many known reasons for this, but frequently it is because the procedure is so difficult that when an installer sees an easier path, he generally takes it.

"Record drawings," or asbuilt surveys, are rarely done following the work is complete. Typically, the conversation goes something similar to this: "Here are the design drawings. Redline any changes that you made."

There is not lots of motivation to do a completely new survey. But in case a design team takes these documents and models them into their computer programs, they're unknowingly creating multiple problems for the contractor on the new job.

We recently took a set of asbuilt documents for a complex project, modeled them and compared them to the point cloud to do a clash detection to determine potential interferences. The outcome was eye opening.

Several pipes, ducts, waterlines or fire lines in the ceiling were in the area shown on the record drawings. If these documents had been used, the MEP contractors would have spent ten times our fee "field fitting" the brand new utilities inside the old.

With the utility and cost of laser scanning, it could be best if you use one on every renovation project. If for nothing else, insurance! Just one field fit will often cost far more compared to the scan itself.

If you scan the surroundings and put the proposed design in to the point cloud, you can tell in just a few momemts where the major interferences will be. We've found conflicts that would have taken upwards of $100,000 to repair if they had to be field-changed during construction. Some were fatal flaws in the required design clearance which could not have been achieved and a completely new design would have had to been submitted.

Scanning to BIM is really a big and extremely important step in surveying. Right now, it is the design software that's trying to meet up with the scanning potential. Already this year, several new programs have come out that are far better at accepting point clouds and computer models, however they still have quite a distance to go.

Not having a design based on a laser scan of the actual environment is a risk that few designers should take. I know I wouldn't desire to tell an owner that there is a construction problem that could have been avoided with a cheap laser scan.

Laser scanning has evolved from a "luxury" to a best practice and it's not a step that any prudent designer should skip.