A 2D or 3D model derived from as-is surveying & scan data representing the asset exactly as it was built. Alternatively defined as-existing, complete with any operational related distortions including heat; gravity; corrosion & settlement
A selection of applicable standards
The standard for data integration, sharing, exchange, and hand-over between computer systems for the Plant Industries. The purpose of ISO 15926 is to provide a common language for computer systems, thereby integrating the information produced by them. Although set up for the process industries with large projects involving many parties, and involving plant operations and maintenance lasting decades, the methodology can be used by other industries willing to set up a proper vocabulary of reference data
The Industry Foundation Classes (IFC) specification is a neutral, non-proprietary data format used to describe, exchange and share information. It is the international standard for building information modelling (BIM) used for sharing and exchanging construction and facility management data across different software application
ASTM E57 standard
The E57 data format is a compact, vendor-neutral format for the storage of point clouds, images and metadata. The data format is specified by ASTM and is documented in the ASTM standard E2807
The Initial Graphics Exchange Specification (IGES) is a vendor-neutral file format for CAD data exchange standard developed in the 1980s capable of exchanging only Geometry and Topology information between different CAD systems. This standard is no longer being developed (latest version being IGES 5.3)
LAS file format
The LAS file format is a public file format for the interchange of 3-dimensional point cloud data between data users. Although developed primarily for exchange of Laser Point Cloud data, this format supports the exchange of any 3-dimensional x,y,z data
STEP / ISO 10303
An ISO standard for the computer-interpretable representation and exchange of product manufacturing information. Known informally as “STEP”, “Standard for the Exchange of Product model data”. ISO 10303 can represent 3D objects in Computer-aided design (CAD) and related information and enables the exchange of data between CAD & CAE, PDM, and other systems across a variety of industries and applications
World Geodetic System (WGS) is a standard for use in cartography, geodesy, and navigation including by GPS. It comprises a standard coordinate system for the Earth, a standard spheroidal reference surface (the datum or reference ellipsoid) for raw altitude data, and a gravitational equipotential surface (the geoid) that defines the nominal sea level. The latest standard is from 1984.
Wrapping a patch-work quilt of freeform NURBS surfaces around scan data, quickly and automatically generating surfaces
See Control Network
Boundary representation models are composed of two parts: topology and geometry (surfaces, curves and points). The main topological items are: faces, edges and vertices. A face is a bounded portion of a surface; an edge is a bounded piece of a curve and a vertex lies at a point
Building Information Modelling (BIM)
A broad term to describe the process of creating a digital model of a building (bridge, tunnel etc). BIM maturity levels are categorised as:
Level 0: Unmanaged in CAD
Level 1: Managed CAD in 2D or 3D.
Level 2: Managed 3D environment with data attached, but created in separate discipline models.
Level 3: Single, online, project model with construction sequencing, cost and lifecycle management information.
Cloud to Cloud Registration
Cloud to Cloud Registration is generally an automated process whereby overlapping point clouds can be registered to each other as consolidated single point cloud of the asset. There are risks with this methodology, particularly for large or complex assets as the integrity of the consolidated point cloud is reliant on all the individual point clouds being of good quality with sufficient overlapping data to allow accurate registration. See alternative via Control Networks
The surveying of larger Sites or Assets begins with the setup of a control network. The control network is established within an Asset or Site coordinate frame to establish the back bone within which all further survey activity is to be referenced. The control network consists of a series of key targets which are carefully placed to place the surveying equipment for detailed capture. Each Target location is recorded using the Total Station; each Target has a unique reference number which is recorded on the instrument. The control points are recorded with sub-millimetre precision to maintain the highest degree of accuracy throughout. Local Networks can then be established with reference to the Control Network
When Plant equipment is closely spaced or complex it can have some degree of Congestion. This can mean that for a small physical area, multiple laser scans or Photogrammetry images are needed to capture sufficient detail for full visualisation or complete modelling
A laser scanner will capture the included objects and their surfaces and also unwanted data points and objects. These can be temporary (vehicles, people, precipitation, materials, etc.) plus measurement errors (poor or misleading readings from reflective surfaces, vibration or movement and other measurement noise, etc.). Included objects may be scanned several times as the scanning equipment is moved around the site. Laser scanners will also capture objects outside the scope of works. Data cleansing is the removal of this unneeded data from the Point Cloud
Data Consolidation/Data Processing
Site consolidation is the process of coordinating and preparing the site data. Scan data is loaded and excessive noise removed; scans are organised into logical volumes. Panoramic image sequences are stitched together to form a single panoramic image of the view location. Any Photogrammetry images are pre-processed to eliminate distortions introduced by the lens – each camera has its own calibration characteristics. Images and scans are registered to the survey network. Panoramic images are combined with the laser scan data to create the colour scan point clouds.
The specification for the modelling (reverse engineering) process should list which engineering disciplines that will be available in the final model. Depending on which are to be included, then the survey (Laser Scanning and/or Photogrammetry) methodology can be adjusted to ensure that all the necessary disciplines are captured into the Point Cloud. This can typically be achieved by varying the settings on the scanner plus the spacing and positioning of the scan locations. A typical list:
- Buried Services
- Cable Ladders & Trays
- Concrete & other Civil Structures
- Electrical assets (inc Junction boxes)
- Hangers & Supports
- Instrument Assets
- Instrument Tubing
- Other Equipment
- Pipe Racks & Supports
- Piping (with minimum size)
- Pressure Vessels
- Stairways, Railings & Ladders
- Steel Structures
Georeferencing is the transformation of a point cloud into a local (facility or plant) coordinate system or a global coordinate system. See WGS84
High-dynamic-range imaging (HDRI or HDR) is a technique used in imaging and photography to reproduce a greater dynamic range of luminosity than is possible with standard digital imaging or photographic techniques. The two main types of HDR images are derived from merging multiple standard dynamic range photographs at differing apertures or acquired using special HDR image sensors which generate HDR images in one pass. The benefit of HDR is that the images are effectively at an even & optimal intensity for human eye (computer vision) readability
Hybrid Surface Model
An IGES or STEP surface that usually combines auto-surfaced features with typical 3D modeling operations. Hybrid models are “dumb” because the data contains no parametric history of the model; it is simply a surface that cannot be intelligently edited. Such models have areas that are not ideally mathematical in nature, and instead are composed of NURBS surface estimates of the scan data.
Combining two different modelling processes to accurately define 3D geometry. This is commonly achieved using NURBS data to define organic (complex) geometry and CAD to recreate prismatic (solid) features.
See ISO 16739
Laser scanning scan the surfaces of objects via a reflected laser beam from a rotating point source to establish an XYZ image of the scanned area. Millions of points are scanned automatically at a very high rates, so each scan location is captured within a few minutes. The data generated is known as a Point Cloud. Post processing via a Registration process combines point clouds from individual scanner locations into one complete Point Cloud. Laser scanners can only scan points that are visible from its location and so objects within the scan may cause a “shadow” obscuring details and other objects which as known as Voids
LIDAR is an alternate more formal name for laser scanning technology and derived from a combination of “Light” and “Radar”
Modelling Terms & Sophistication
There isn’t a standard definition for the sophistication of the information as opposed to dimensional accuracy of a 3D model in the Plant industries, although many companies use similar terms. The table below is a comparison of the various levels of sophistication
While Point Clouds can be directly rendered and inspected usually point clouds themselves are generally not directly usable in most 3D applications, and therefore usually converted to differing model types with varied levels of sophistication in the information both in the Geometry and the data of the model element
|Low||Cloud; Raw scan or image data|
|Nurbs Surface; B-Rep|
|ShrinkWrap; Colour Draped|
|Dumb; Primitive; Dumb Solid; Geometry Only|
Multi-view Stereo (MVS)
Mathematical representations of 2- or 3-dimensional objects, which can be standard shapes (such as a cone) or free-form shapes (such as a car). NURBS are used in computer graphics and the CAD/CAM industry and have come to be regarded as a standard way to create and represent complex objects
A (digital) photograph geometrically corrected such that the scale is uniform and has the same lack of distortion as a map. The original image has been adjusted for relief, lens distortion and camera tilt. An uncorrected perspective photograph Orthophotograph cannot be used to measure true distances
An image encompassing a field of view which extends beyond a traditional photographic image. Usually encompasses 180 or 360 degrees across the horizontal axis. Specialist cameras can capture panoramic images or multiple images can be combined to make a composite panoramic image.
Uses parameters to define a model (dimensions, for example). Examples of parameters are: length, width, diameter, material density, orientation etc. The parameter may be modified later, and the model will update to reflect the modification
Does not contain any intelligent data, e.g. temperature and pressure data or related piping. Can be modelled with different levels of detail, beginning with the simple representation of areas as envelopes (blocks) and ranging to fully detailed models. Combinations of different levels of detail are quite common. Dumb IGES – a term used to refer to any IGES, STEP or another surface file format. Technically a mathematical model, it is considered “dumb” because the data contains no parametric history of the model; it is simply a surface that cannot be intelligently edited. Ex: If a cylinder is modelled in 3D and exported in IGES file format, the cylinder cannot be edited by changing its radius or extrusion length
See Colour Draping
The science of making measurements from photographs, especially for recovering the exact positions of surface points in close-range Stereo Photogrammetry a calibrated high resolution (digital) camera takes multiple images close to the object (hand-held or on a tripod). For 3D scanning & surveying, coded targets and reference markers in the picture frame serve as anchor points which images can be aligned to. Computer based algorithms then determine surface points from 2 or more referenced images to generate an extremely accurate model or point cloud. See Structure from Motion (SfM)
A point cloud is the computer visualization of the XYZ coordinates that describe a physical object or environment. Each point represents an actual point on the object or in the environment, and collectively the Point Cloud describes its shape and measurements. Points are captured individually via a 3D laser scanner. The capture of multiple Point Sets from different perspectives that can be merged into a Point Cloud via Data Cleansing & Registration
Registered Point Cloud
This is a single or aggregated point cloud that consists of several scans and has been transformed into an integrated and higher-level coordinate system. The coordinate system may be global or local (i.e. a plant).
See Unified Point Cloud
Scanning is normally done from several locations in order to scan an object completely. This produces several point cloud sets (individual scans) in a localized coordinate system. The individual point clouds are combined into one large complete Point Cloud in one integrated coordinate system. This step is called registration
A graphical representation of a (CAD) computer model. It is often used to describe the visual output of CAD and Modeling software. By rendering a computer model, characteristics & effects can be added to the its surfaces plus motion & fly-throughs
See Point Cloud.
Scan resolution is specified by the number of points in a laser scanner’s field of view or by the distance between the measured points at a specific distance from the laser scanner. Because laser scans effectivley originate at a point source and the equipment is limited by rotational accuracy, more distant objects are scanned with reduced resolution and may miss important detail. This requires the numnber of scan locations to be matched to the size of asset to be surveyed and the detail/accuracy required for the 3D model
Generic term for Laser Scanners
Shrink Wrap Surface Model
A Shrinkwrap model is based on the external surfaces of the model element(s) and captures the skin, or outer shape, of the source model. Similar to how plastic shrinkwrap shrinks down onto a part being wrapped
Offset Services proprietary technology that combines high resolution 360-degree panoramic image capture with millimetre accurate laser scan data plus intelligent CAD model into a lightweight, browser based viewing environment. This allows a fast and information-rich Google Street View type experience throughout the virtualised as-built intelligent asset
The STL format was originally designed to allow data move between CAD programs and stereolithography equipment. ‘STL’ stands for Surface Tesselation Language or ‘STereoLithography file’ or ‘Standard Transform Language file’ Now widely used in Reverse Engineering
Overlapping pairs of images that share a common subject and have been registered into a known coordinate system allowing accurate 3D measurements to be taken by triangulating corresponding image positions.
Similar to Stereophotogrammetry, the SfM (MVS) technique requires multiple, overlapping photographs as input to feature extraction and 3-D reconstruction algorithms but without the requirement to control the camera orientation or reference targets
The process of defining or creating a surface on a CAD model. Typically refers to converting a polygonal representation of an object to a NURBS or other mathematical representation. It is the process of converting physical based 3D data to mathematical based
Tags are attached to both movable and fixed assets as an identfication mechanism for lifecycle management, inventory control, and Computerised Maintenance Management Systems (CMMS) etc.
Targets are precisely defined points needed for registration. Targets are used to combine and transfer individual scans to a higher-level coordinate system. The shape and appearance of targets vary. Targets may, for instance, be black and white or spherical. Targets are installed in the scan area during laser scanning and are subsequently captured by tachymetry. Spheres constitute spurious objects in a point cloud. This can cause problems, especially for automatic collision analyse See Control Network
Generally refers to filling a surface plane or surface with shapes that create no gaps or holes. In 3D scanning, this concept applies to wrapping a mesh around a CAD body. A tessellation is a gap-less, repeating pattern of non-overlapping figures across a surface. Any shape can be used; STL format uses triangles. This triangular mesh is most often derived from the surface – and only the surface – of a 3D CAD designed object.
Graphic representation of the surface features of an asset or surfaces indicating their relative positions and elevations.
Total Station is a theodolite integrated with an electronic distance meter (EDM) to read slope & distances from the instrument to a particular point. These are used to construct a Control Network & Local Networks
Using trigonometric functions to calculate measurements, used in certain types of 3D laser scanners plus stereophotogrammetry to determine point locations based on transmission and reflection positions of the laser beam. In 3D modeling, triangulation also refers to the generation of triangles out of point cloud data in creating 3D surface.
A combined set of point clouds which has had duplication reduced
A laser scan or photo imaging originates from a point source. The missing data area in the resultant Point Cloud is the “shadowed” area of the laser scan or camera view caused by an object in the foreground is called a Void. Also known as occluded areas