Future of Scan-to-BIM Process with ML Innovations | Blogs | nCircle Tech
Spread the word

Numerous advantages of implementing BIM, like lesser chances of raising more RFIs and frequent change orders, are making it easier for stakeholders to cut down on construction time and cost. However, the dynamic integration of BIM with facility management (FM) practices demands having a detailed as-built building model. Integrated BIM-FM systems offer crucial advantages such as better visualization of the areas and facilities involved within a thorough virtual environment.

Now, the more complex a building is, the more trades are involved. For example, a medical building is required to have more facilities than a school building. The more facilities get involved, the more complex management and maintenance of these facilities become. Usually, coordination issues related to MEP/ MEPF elements possibly get detected within the BIM Model but are worked on at the Jobsite, and those modifications are not typically reflected in the BIM, which makes it hard to manage facilities throughout the project lifecycle.

For better, accurate, and comprehensive as-built information companies are increasingly relying on 3D point cloud data that are used for creating Point cloud as-built Revit models. The process of designing an as-built model from point cloud data is referred to as Scan-to-BIM.

 

Technologies Behind Error-Free As-Built Data Acquisition

As-built data retrieval is the method of obtaining the accurate shape and structure of an object in the point cloud format. Prevailing as-built information acquisition processes essentially include manually performed on-site surveys. Manual surveying is the most used as-built information retrieval approach even today in many construction projects. However, it’s not recommended for large projects. Manual surveying is time-consuming, chances of human error are common, which makes reliable decision-making really challenging.

At special use building projects, as-built data are getting gathered by adopting two non-contact spatial survey technologies, based on photo/video technologies and physical laser scanning

Here, these two as-built data acquiring technologies:

  1. Photo/video-grammetry
  2. Terrestrial laser scanning
  3. Photo/Video-Grammerty

Photo/video-grammetry has made it possible to get accurate 3D reconstruction of an object or construction element that is either under construction or in operation. The basic principle of photo/video-grammetry is triangulation, in which a point in space is reconstructed from a pair of mathematically converging lines from 2D positions of the point in different representations or images. Once camera orientation and calibration are measured, collected images or video frames are utilized to determine the geometric features of an object. After performing this the object can be reconstructed in three dimensions by applying image-matching techniques.

 

Terrestrial Laser Scanning

With Terrestrial laser scanning or Laser Distance and Ranging (LADAR), 3D coordinates can be obtained directly by scanning the surface of a target object by using advanced laser scanners.

Depending on how advanced sensors have been used, the characteristics of the surface being scanned, or the distance of the target surface from the laser scanner, each measurement can be precise to a few centimeters or less than a millimeter. TLS enables collecting specific 3D point cloud data in a short amount of time. For tempting reasons like this, it is extensively used in the AEC-FM industry to scan different trades and elements of construction projects.

 

So, What Exactly is Point Cloud Data?

Without getting much into the nitty-gritty of Scan-to-BIM workflow, let’s take a moment to understand what Point Cloud data is.

On construction sites, large laser (Scan-to-BIM) scanners are used to capture detailed data providing exact and minute information about every nook and cranny on a job site. This data gets collected in the form of cloud data which consists of millions of 3D points. Scan-to-BIM scanners can collect detailed point cloud data and these datasets can be used to develop digital 3D models of the scanned environment.

The key to create a realistic and detailed model is having accurate measurements to work on. With Scan-to-BIM scanners or 3D Laser scanners, accurate measurements can be obtained. Scan-to-BIM process allows a data-rich, well-connected, and easily constructible workflow.

While we can dive deep into the world of point cloud data and what it can do. However, in this article, we are going to focus on the most important advantage of scan data, which is to offer a remarkably detailed digital record of a construction or space.

Ø  Scan-to-BIM Workflow

In a Scan-to-BIM workflow, a laser scanner is used to get a precise 3D scan of a target object. The Point cloud data is then imported into a three-dimensional modeling environment to build detailed as-built models.

Ø  Scan-to-BIM or Point Cloud Modeling Services

Scan-to-BIM services providers utilize software to create accurate and efficient Point cloud as-built Revit models that can be used for further project development and refurbishment of the infrastructure.

nCircle Tech offers efficient and cost-effective Scan-to-BIM services. Our in-house experienced BIM team leverages various Scan-to-BIM tools or plugins to help in converting the point cloud data into the detailed BIM model. We create different software Plugins that enable users to publish their models directly from one software environment to another without leaving the previous one.

Many software development companies have implemented semi-automated tools into existing Scan-to-BIM software. For example, one such software is Cyclone developed by Leica Geosystems, Ag, RealWorks developed by Trimble Navigation, Ltd., and EdgeWise3D developed by ClearEdge3D.

Ø  Leica Geosystems

Leica Geosystems developed software including various functions that allow 3D modeling from Scan-to-BIM. The latest version of Leica Cyclone offers users an interface including tools for extensive pipe modeling with different functions like automatic pipe finder, cylinder fitting, and model creation from point clouds.

Leica Geosystems also has released numerous plug-in tools for component-specific modeling from 3D Scan-to-BIM data. By using these plug-in tools, it is now attainable to import 3D point clouds inside Revit, AutoCAD, and Microstation.

Ø  Trimble Navigation:

Trimble RealWorks is a point-cloud processing software that offers 3D modeling features. Trimble RealWorks Advanced-Modeler module allows creating models from point cloud data, especially of objects like columns, beams, walls, and others.

Trimble RealWorks Advanced-Plant module is provided with modeling functions, particularly ones related to the power, process, plant, and associated environments.

Another Trimble product to create 3D models from point cloud data is Trimble SketchUp.

The Trimble Scan Explorer Extension allows importing 3D point-cloud data into Trimble SketchUp seamlessly to create models from 3D scan data.

Trimble SketchUp possesses an automated plane extraction tool to improve modeling efficiency, especially when creating building interiors and facades.

Ø  ClearEdge3D – EdgeWise

EdgeWise is an as-built modeling platform that applies superior algorithms to automate tiresome scan-to-BIM workflows. Offers Scan-to-BIM solutions by creating models of structural elements, as-built pipes, ducts, conduit, and cable trays from point cloud data faster and more accurately. EdgeWise provides a robust algorithm that can manage large 3D Scan-to-BIM data.

 

Limitation of Popular Scan to BIM Software

The main limitation of the existing scan to BIM software is consuming a significant amount of time to import scanned data into a 3D modeling environment. Usually, it takes days of time to process the scanned data with conventional Scan to BIM software. 

For example, for a 1000 Sq. ft space, creating LOD 300 BIM model out of point cloud data can take upto 15 days time. As point cloud data is humongous and 10 times larger than the size of the BIM model, it becomes quite hectic for engineers to work with the huge amount of data. Getting the point cloud data of an existing structure using a 3D scanner isn’t a big deal. However, the generated data can be unstructured and working on heavily unstructured data poses a challenge, quite hard to overcome.

In traditional scan to BIM workflow, a lot of manual intervention is required which isn’t ideal when you need a faster output. While point cloud scanning is increasing, limitations are still seen lurking around the industry. 

 

ML Based Scan to BIM Solutions Emerging As the Game Changer 

For modelers and engineers who are tired of tedious manual work, nCircle Tech’s ML-based solution can prove to be a saviour. Our Machine Learning based Scan to BIM solution automates traditional scan to BIM workflow by spontaneously performing 80% of the conversion work in less time. Faster conversion means better and faster BIM model generation. Post conversion, modelers can create an accurate BIM model on LOD*1100/200/300/400/500 which again can get converted into revit2, AutoCAD3.

Our Automated ML Based Scan-to-BIM Process allows users to follow the usual Scan-to-BIM workflow while reducing the cognitive load needed for visualization. You may wonder how exactly we offer ML Based Scan-to-BIM service. Here is how?

While loading the 3D point cloud model into the software of your choice, our in-house team processes the 3D point cloud model through the Scan-to-BIM framework. This process gives users the same 3D model with the annotations in the “Viewing mode”.

Our machine learning Scan-to-BIM process results in a framework that identifies the various elements present in the point cloud model and classifies them with varying color codes.

 

Advantages of Machine Learning Scan-to-BIM Process

At nCircle Tech, we have developed an ML-powered Scan-to-BIM service. By providing this ML-powered scan-to-BIM service we help clients to identify BIM objects from scanned data and make the complex process of Scan-to-BIM a level more convenient. The entire Scan-to-BIM process becomes smarter using the power of Machine Learning.

  • Quicker turnaround time; we use an ML Based Scan-to-BIM algorithm to derive features from Point cloud data
  • The Scan-to-BIM using ML output can be read into any BIM software
  • The Scan-to-BIM using ML algorithm can work with a huge amount of data at once.
  • While utilizing ML Based Scan-to-BIM, frequent errors can be avoided, which eventually reduces cost.
  • Limited people need to visit the project to gather data using Scan-to-BIM scanners

nCircle Tech works as Software Application Plugin developers to modify the existing modeling and designing process. 

 

The Road Ahead!

Our ever-evolving ML-based Scan to BIM software solutions efficiently convert point cloud data into a 3D BIM model in lesser time than traditional scan to BIM solutions and produce more accurate output. Using ML engine, our scan to BIM solution generates more structured data which can further be utilised to use accurate BIM models.

Sounds promising right? Didn’t we just make Scan-to-BIM way more simpler? What’s more? Well, this standard ML-powered Scan-to-BIM framework can be customized and implemented as per your specific workflows.

nCircle Tech aims to inspire passionate innovators to create impactful ML-supported software for more agile Scan-to-BIM model outputs on desktop, mobile, and cloud platforms. Our decade-long domain expertise and impressive track record empower our clientele to reduce time-to-market, crossing business milestones, and going beyond them.


Spread the word