Between the Poles

US 3D elevation project as a model for a national infrastructure map

A number of years ago an initiative was funded under the ARRA economic stimulus package to initiate what was then called LiDAR for the Nation.  This has resulted in a USGS initiative called 3D Elevation Program (3DEP) whose goal over an eight year period is to develop a 3D elevation map with horizontal resolution of a meter and vertical resolution of 10 cm for the entire U.S. and territories.  This entails LiDAR scanning the contiguous continental U.S. and Hawaii and interferometric synthetic radar over Alaska.  It also includes bathymetric data for the U.S. coastline.  To date 48% of the U.S. has been covered or scanning is underway.  An inventory of these projects is available.

At GIS in the Rockies this year Carol Lydic of the USGS gave an overview of this program and the most recent initiatives to broaden the partnerships that are instrumental in creating the national 3D elevation map.  The program is based on partnerships with federal, state, local, and tribal government agencies, academic institutions and the private sector.

In 2012 a study was completed that estimated the economic value to 27 different sectors of accurate 3D data.  The top five included in this study (National Enhanced Elevation Assessment) are

• Flood risk management                                  $295m to  $502m
• Infrastructure and construction management $206m to  $942m
• Natural resources conservation                     $159m to  $335m
• Agriculture and precision farming                  $122m to $2,011m
• Water supply and quality                                  $85m to  $156m

The total economic benefits for all sectors was estimated to be between $1.8B to $13B.   I find this to be a very conservative estimate. 

A number of years ago I remember hearing about an outstanding program by the state of North Carolina in cooperation with federal agencies that involved a statewide LiDAR scan and the creation of a database of all structures with a footprint of over 800 square feet that allowed estimates to be made of the economic impact of any type of flood anywhere in the state or of a rise in sea level along North Carolina's coastline.  This is of inestimable value in assessing any mitigation project.

In the infrastructure and construction sector, one of the largest contributors to the cost of a civil construction project such as a highway or a railway, is the amount of earth that has to be moved which can be derived from a 3D terrain model.  Having a reliable digital terrain model is of great value to any civil project for assessing alternative right of ways during the planning stage and could also be used during the design and construction phases especially if supplemented by information about the soil and geological structure of the subsurface and location of underground utilities.

I found the 3DEP project particularly interesting because it represents a potential model for creating a national infrastructure map including underground infrastructure.  Federal organizations that could be involved include USGS, National Map, US DOT, FERC, and NERC.  Partners might include NARUC (National Association of Regulatory Utility Commissioners), CGA (Common Ground Alliance), state and municipal DOTs, AGC (Associated General Contractors of America), utility and telecom network operators, general contractors, tribal government agencies, and academic institutions.

Electic power utility uses reality capture to rapidly develop grid data model

Duquesne Light is typical of many of North American small to medium electric power utilities.  It does not have an outage management system  or an electrical data model which means it is difficult to connect customers to transformers and to determine upstream devices when customers report power problems.  It has become clear to Duquesne Light that an electric data model is essential for moving forward on modernizing their grid. At Distributech 2018 Jim Karcher, Manager, Operation Technology Projects at Duquesne Light Co. (DLC), gave an overview of a remarkable project that compared different approaches including reality capture and a traditional foot survey for developing an electric grid data model.

Duquesne Light still uses paper tickets for tracking outages and other problems.  Currently Duquesne Light's circuit maps are AutoCAD drawings.  These drawings are designed to be used by electrical engineers and are not spatially correct.  The reliance on AutoCAD drawings means that  tracing from transformers to customers or from customers to controlling devices is a manual process.  However, Duquesne Light is moving to modernize its grid.  It  has implemented distribution SCADA (DSCADA) and has over 1100 DSCADA controlled field sectionalizers, about 400 DSCADA controlled field capacitors and is implementing an AMI deployment which will be completed in 2019.  And it plans to deploy and Advanced Distribution Management Systems (ADMS) including outage management (OMS), DMS and DSCADA.  But the ADMS will require developing a complete distribution electrical data model.

A quarter of a century ago there were two ways to do this.  Either walk the entire network and plot it on a USGS map or attempt to conflate the AutoCAD drawings onto a USGS basemap.  Today's technology advances have created new alternatives.  One of these is to use modern reality capture, typically LiDAR and photo and infrared cameras with a GPS which can be mounted on a truck. 

To compare alternative ways to create a grid data model Duquesne Light decided to run a pilot on about four of its circuits comprising a 100 miles of its distribution grid.  Quantitative criteria were defined to allow comparison of cost, schedule, and electric data model accuracy of different approaches.  Electric data model accuracy included location accuracy, phasing, ability to trace customers to transformers, asset inventory and attribution.  The data model had to be captured in a form that was compatible with Duquesne Light's GIS.

Reality capture was done with a pickup truck with side mounted photo and IR cameras and LiDAR laser scanners.  Pictures were taken very second which enabled a 270 degree picture of each pole.  Location accuracy was about a meter.  To drive the entire 100m miles of distribution network required about a week.

When the model captured by reality capture was compare to the result of a traditional survey (on foot with a GPS), only one discrepancy was found that could be attributed to the reality capture approach.  Comparison with other approaches revealed that the reality capture data provided the most geospatially accurate asset location data.  The reality capture survey only required two Duquesne Light employees in the field for a week and at no time during the survey did they have to leave the truck - meaning that the survey was completed rapidly and safely.  Actual development of the data model was completed in the office without the necessity of going to the field for quality checks because the imagery provided all the data that was required to create and verify the data model.

Duquesne Light found several significant advantages of the reality capture approach.  The first was asset accuracy - both location and asset attributes.   Reality capture provided a visual record that could be checked in the office obviating the need to go to the field.  The captured imagery could be used for other purposes.  It could be used for a joint-use audit.  The LiDAR data allowed distances between poles and cables to buildings and structures to be accurately measured to ensure regulatory clearances and to verify third party attachment heights.  The infrared imagery could be used to detect hot spots.

This is critically important because it shows that reality capture is a feasible alternative to a traditional foot survey.  Duquesne Light has shown, based on a quantitative comparison of alternatives for developing a data model including a traditional foot survey, that reality capture provides a cost-effective and safe way to rapidly develop an accurate electric data model.  It also concluded that reality capture also provides significant additional benefits.

Fully automated classification of road and highway features for autonomous vehicles

Point clouds are fine for visualization to provide context, but to be really useful they need to be classified.  Feature extraction, as the process is often referred to,  is currently a mostly manual process that has been used for many years in the remote sensing sector. A trained person visually inspects the point cloud and graphically selects features which are identified by type of feature, sign, light pole, lane marking, and so on. 

Automated classification of point clouds

Image processing of 3D point clouds has advanced to the point where it has been possible to semi-automate this tedious feature extraction process.  The reliability of the automation process depends on the point density of point cloud, often expressed as points per square meter or per square centimeter.  Point density depends on the technology used to capture the point cloud. Terrestrial lidar, for example, typically captures points at the rate of hundreds of thousands to millions of point per second and is capable of millimeter to centimeter precision. Phodar generates lower point densities and is less precise. Aerial lidar captures many points per second but with less precision because it is covering much greater areas. Single photon lidar is a new technology that can collect 6 million points per second which means that planes can fly quite high, cover very broad areas and still collect 12 - 30 points per square meter.

At the Year in Infrastructure 2017 (YII2017) conference in Singapore, Sharad Oberoi of Sanborn gave an insightful overview of the state of the art for automated feature extraction from combined aerial and mobile scans of highways and adjacent areas.  Different types of laser scanners are appropriate for capturing different types of features. Terrestrial and mobile laser scanners can capture point clouda that are suitable for extracting curbs, gutter drains, traffic signals, road and local terrain information, parking meters, walls, obstructions in the right of way, manholes, sidewalks, overhead clearances, lights, utility wires and connections, garbage cans, benches, fences, guardrails and barriers, retaining walls, and line of sight vegetation obstructions. Aerial lidar will only be able to get gutter drains, road and local terrain information, walls, obstructions in the right of way, and sidewalks with limited ability to get curbs, lights, utility wires and connections, retaining walls, and vegetation.  Lidar does not do well for metal structures such as communications towers and fir these a combination of lidar and phodar is preferred.

Sanborn has developed feature extraction technology that extracts transportation objects from combined aerial and mobile data sets in a semi-automated manner. The mobile and aerial data sets meet the accuracy and information content required for geospatial information for mapping applications as well as the additional information that can be mined for potential asset inventory and infrastructure information content. Mobile technology allows for the low-risk, rapid collection of geospatial information, limiting safety impacts to workers. The aerial LiDAR allows for the rapid collection of elevation information for detailed surface modeling as well as feature extraction. Both technologies collect enormous amounts of point-cloud data that can be merged using the Terrascan capability in Bentley MicroStation. The combined data set is then passed through the automated feature extraction modules to identify transportation objects.  Sanborn feature extraction modules achieve a high proportion of correct feature extraction for road and highway scans.

In the future Sharad expects that it will be possible to improve data collection efficiency by using terrestrial robots or autonomous UAVs equipped with SLAM technology to enable them to operate in GPS-denied areas such as under bridges and in tunnels.  Sanborn has been working with deep learning to improve the automated extraction process.  So far this has yielded mixed results, but Sharad expressed optimism that with AI full automation of the classification process will be achieved.

High precision maps for autonomous vehicles

An application of this technology is high precision maps for the autonomous vehicle market.

Sanborn develops high precision maps for the autonomous vehicle market.  Autonomous vehicles will require high precision maps which contain significantly more detailed information and true-ground-absolute accuracy than current road maps. Sanborn has developed proprietary mapping technology that leverages aerial imagery, aerial lidar data, and mobile lidar data to create standardized, high-precision 3D base-maps focusing specifically on the self-driving vehicle market. These maps include detailed inventories of all stationary physical assets related to roadways such as road lanes, road edges, shoulders, dividers, traffic signals, signage, paint markings, poles, and all other critical data needed for the safe navigation of roadways and intersections by autonomous vehicles.

Geospatial technology captures wall instability and collapse in an open pit mine

One of the most exciting presentations at the SA Geotech conference in Johannesburg was a presentation by Jason Richards of Maptek.  He talked about using geospatial technology to map in space and time the instability and ultimate collapse of a wall of the Kanmantoo Copper Mine in South Australia.  This is an open pit copper mine that was in operation 1970-1976 then closed but has reopened after significant additional deposits of copper and gold were confirmed.   The copper and gold deposits are mined using well established open-pit excavation techniques including drilling, exploring, blasting, loading and hauling. Two operational pits were dug at the mine site to a depth of 270m for extracting the reserves.  The larger pit (Giant Pit) spans up to 1.2km long and 600m wide. Hydraulic excavators and highway trucks are used to transport the ore to the processing plant.  The roads used by the trucks wind along the walls of the mine.  The surrounding rock is brittle and liable to fracture.  Collapse of a wall has occurred in the past and is a serious hazard for the trucks carrying the ore out of the mine for processing.

Unexpected geological conditions occur in the east wall of the Kanmantoo Giant Pit resulting in threatening wall movements.  A Maptek Sentry system was put in place to map wall movements to monitor the stability of the wall.  The Sentry system combines laser scanning and photography and can monitor multiple areas within a scene in real time without the need for targets or reference points. Movement can be detected down to 1 millimeter per hour. Heatmaps colored by displacement or velocity provide an overview of movement in real time.  Animations of time changes can be used to show morphology of terrain changes over time. Time lapse capability allows users to see changes and use this information to predict future movement.

In the Giant Pit a Sentry system was set up to monitor a 550 meter wide section of the brittle east wall of the Giant Pit at Kanmantoo.  The wall was scanned initially every 23 minutes, but after significant continuing movement was detected scans of a more focussed area were made every 7 minutes.  Movements were detected that were directly related to blasting with sharp jumps in displacement that usually stopped moving after an hour or so.  But after one blasting sequence a 364 tonne block initially jumped 30mm with the blast.  Two hours later it accelerated again, ultimately failing 4 hours after the initial blast.  The Maptek system was able to capture the movement, raise an alarm and visualize the movement with heat maps and video.  Jason showed a time lapse video of the movement and ultimate collapse of the wall which is one of most impressive videos of the progress of a geological event I have seen.

This is another example of how geospatial technology, in this case laser scanners combined with photography with real-time geoprocessing, is able to provide solutions in a wide range of industries.

Perceptality: new laser scanning applications potentially transforming construction

At HxGNLive Juergen Dold, President of Hexagon Geosystems which is responsible for the Leica Geosystems brand, gave a talk about perception and reality.  Perception is how we see the world, but perceptions are subjective and are not always reliable.  Reality can be quite different.  If you have ever worn a pedometer,  you will know exactly what he is referring to - the actual number of steps you walked or ran is often quite different from what you think you did.  Until now measuring things in construction and other industries has been so laborious that it is frequently not done.  Decisions may end up based on the "biggest salary's" opinion.  With laser scanning, photogrammetry, UAVs and other technologies measuring things has become several orders of magnitude easier and we can replace some of our perceptions with reality - Juergen called it fusing reality and perception - and coined a new term for it perceptality.  Since as Peter Drucker said, you can't manage what you don't measure, increasing the reality domain at the expense of someone's perception has the potential to dramatically change construction.

Juergen provided several interesting real world examples of how this translates into practice.

1 Vegetation management for transmission lines - inspecting transmission lines for vegetation encroachment in North America is mandated by NERC.  This is typically an expensive, slow process involving a helicopter carrying a LiDAR camera.  To capture the necessary density of points to be able to identify cables, pylons, and different types of vegetation requires flying at low altitude.  With the latest single photon laser scanner (Leica SPL100), it is possible to collect 80 points per square meter (m²) flying at an altitude of 2,500 meters (about 8,000 ft) as opposed to 3 points per m² with traditional linear laser scanning.

2 Urban mapping - urban mapping can also be time consuming and expensive, but with multiple cameras including LiDAR, photography, and near infrared it is possible to capture a more realistic 3D representation faster.

3 Scanning industrial plants - Cepsa (Compañía Española de Petróleos, S.A.U.) is a Spanish energy company who are scanning their chemical plants which requires thousands of scans.  A major software advance (Leica Cyclone) allows them to automate the registration process saving a lot of time.

4 Forensics - scanners in backpacks (Pegasus Backpack) or handheld scanners make it possible to walk through the scene of an accident or crime and capture trusted data without disturbing the site.  The data can be used in scientific forensic analysis and to visualize the results of the analysis in court.

5 Digitalizing AEC - The construction industry is one of the most inefficient industries.  $10 trillion is spent on construction goods and services every year. But whereas the average worker adds $37 of value for every hour worked, a construction worker adds only $25 per hour worked. Construction productivity has stagnated over the past decades in many of the world's advanced economies.  To attract the private financing that is necessary to build and maintain the world's infrastructure will require improving productivity.  Digitization can play an important role in this process or even be disruptive (transformative) as it has been in other industries. Innovative firms have increased productivity by 50 to 60 percent by infusing digital technology, new materials, and advanced automation, reskilling the workforce, changing business practices and pushing government to change regulations.

Our existing stock of structures will be with us for many years to come.  Maintaining this infrastructure is currently inefficient and expensive.  New technology can dramatically change this by making renovation more efficient and affordable.  A first step is democratizing scanning.  I have blogged about the deluge of handheld, low cost scanners that are now available.  Many see handheld scanners as potentially revolutionizing the construction sector.  Mantis Vision has offered a handheld scanner F5 for professionals for several years.  Originally targetted on the real estate industry, the Matterport Pro allows you to scan a house in an hour or less, upload the captured data to the cloud (Amazon Web Services) and a couple of hours later you will have a fully-rendered model that can be used for walkthroughs with potential buyers.  The Faro Freestyle is targetted on the architecture, accident reconstruction, civil engineering, construction, facility management, and industrial manufacturing sectors. 

I have blogged previously about the Leica Geosystems BLK360 and Cyclone Register 360  which together appear to be a game changer for professionals in construction.  It is a bit of a stretch to call it a handheld scanner, but it weighs only 2.2 lbs (1 kg) and is incredibly small 6.5 by 4 inches.  Unlike other "handheld" devices it is a full laser scanner and imager - it captures 360,000 points/second with a range of 0.6 - 60 meters with millimeter precision. It has three cameras, laser scanner, photo, and near infrared.  It's one button operation redefines simplicity.  It can capture a full 360 degree scan in less than 3 minutes.  Cyclone Register 360 allows automated registration of scans from many devices to be merged with "one-button" simplicity.

The other half of the laser scanning revolution in construction

I have already blogged about the Leica BLK360 whose size, capabilities, one button operation and price point promise to revolutionize professional scanning for construction.  Anyone who has used a Leica scanner is familiar with Cyclone software and, in particular, with  Cyclone Register which is used to register point clouds when carrying out scanning at construction sites.‎ Professional surveyors who have been using Leica P30/P40 and other scanners have learned to navigate the complex Cyclone Register workflow. But what about the people for whom the BLK360 with its one button operation is opening the door to scanning for the first time ? Having to learn Cyclone Register could be a frustrating and potentially showstopping experience.

This is where Cyclone Register 360 comes in. At HxGNLive this morning Greg Walsh, who leads the software design group that developed Register 360, presented an overview of the the thinking, design objectives and implementation behind Register 360, which is just as important as the hardware in BLK360 in revolutionizing the professional use of laser scanning in construction.

Greg described a completely new way to register point clouds from many scans, relate scans to each other, optimize, generate deliverables and report estimated errors for scans involving multiple BLK360s deployed on different floors and in different areas of a construction project. To do this simplified workflow requires new terminology.  This will make things easier for people new to laser scanning, but will require learning a new perspective for people with scanning experience.  Register 360 is not intended to replace Cyclone Register, so folks who are familiar with the traditional Register can continue to use it.

For professionals for whom the BLK360 is their first scanning experience, Register 360 is intended to dramatically simplify the registration experience even for scans of large complex construction sites. Three "big red buttons" analogous to the single button on the BLK360 should enable anyone to register scans, optimize and create finished scans for visualization and analysis using other products such as Revit, TruView and Jetstream.  Underneath the covers there is heavy-duty processing going on - we are dealing with billions of points after all, but the software is designed to do the heavy-lifting while you are doing other things such as getting coffee.

Currently Register 360 is intended for terrestrial scanning using BLK360s on tripods. But it is planned to extend it to mobile and other types of laser scanning.‎ For those people who read manuals it comes with a tutorial, but the implication of three red buttons is that the tutorial is optional.  The BLK360 is not a consumer device, but with Register 360 it should dramatically extend the professional user community beyond surveyors.

Practical BIM/laser-scan round tripping on construction projects

At HxGNLive in Las Vegas this week a very interesting presentation with practical implications for anyone involved in construction, whether buildings or horizontal infrastructure was provided by Stanley Lawrence, Reid Flamm, and Michael Moore of Datum Tech Solutions, a Hexagon partner who focus on bridging the divide between the model and the field for owners.  They described a very practical approach applying laser scanners to ensure that the owner is reassured during construction that what is actually being built is what was designed .  Since this is a Hexagon event they reported used Leica ScanStation P30/P40 scanners with Cyclone software.  The model is designed in Revit.  I expect the process is general and could be used with other hardware and software.

 

The detailed process involves setting up the scanner and capturing control points and then performing scans.  The point clouds are imported into Cyclone and then exported to a PC.  The data is registered and aligned and a ModelSpace coordinate system is setup for Revit.  The scanned data and coordinate system are imported into Revit using CloudWorx.  The scanned data is used to manually create a level of detail (LoD) 350 Revit model.  To verify that their model accurately reflects what was scanned, they take their scanned data back out into the field and verify it with a Total Station.  Then they use Autodesk Point Layout (APL) to compare what has been built to what was designed and look for divergences.

 

They report that this process is very efficient so they are able to carry out scans frequently during construction to catch problems early.  For example, they will scan right after pouring concrete for floors or after installing electrical wiring, plumbing or HVAC for a floor.  In addition to keeping the owner in the loop during construction, at the end of construction they are able to provide the owner with scans of the final building.  They also can provide scans from during construction to provide digital "as-builts" of hidden infrastructure such as electrical wiring, HVAC and plumbing.  Owners find that this information is useful for facilities management during operations and maintenance.

Simplifying the integration of reality capture and asset management

The construction industry in the UK, as in many of the world's advanced economies, suffers from low productivity compared to other industries, a situation which hasn't improved in 25 years according to Simon Rawlinson of the Construction Leadership Council.  Various types of laser scanners, software that creates point clouds from photos, and handheld scanners have the potential to dramatically improve productivity, but users of these devices in the construction and asset management industries are finding that these devices are creating a major data management problem which has to be addressed to enable the technology to achieve its full potential.  At Geo Business 2017 in London I had a chance to chat with Pascal Martinez, Director Business Development at Bentley Systems about Bentley's vision for converging reality capture, reality modeling, and asset management assist in improving productivity in the construction and maintenance of building and infrastructure assets.

Pascal was originally with Acute 3D which was acquired by Bentley two years ago. Acute 3D, now called ContextCapture, was originally targeted at creating point clouds from photos captured with an ordinary digital camera.  The latest version of ContextCapture is able to integrate point clouds from LiDAR and photo cameras to generate a common 3D mesh using imagery from both sources.  

Bentley's vision for the evolution of reality capture is to be able to integrate data from multiple sources, both 2D and 3D, to create a common 3D mesh, classify it by associating areas of the mesh with existing data from other sources stored in databases such as Bentley's Assetwise and Projectwise and then link objects in the mesh to information about the specific assets stored in the databases.

At the conference the model of what already exists or what has actually been constructed was frequently referred to as the digital twin of a building or other type of infrastructure (previously BIM).  Associating classified elements of 3D meshes from reality capture with vector elements in the digital twin has a number of applications.  In the case of a new building or new infrastructure it enables monitoring compliance with design during construction, even on a day by day basis, to catch issues early.

Being able to automatically associate areas of a 3D mesh with buildings, facilities and equipment in plants, and linear assets such as transmission lines, railways, and roads would have a dramatic impact not only on the efficiency of construction monitoring but also on asset management by reducing the time require to scan all or part of a facility and compare the captured reality with the digital as-builts. Pascal showed a very simple, but very cool, way to do this now, by placing a QR code on each piece of equipment which then can be clicked on in the mesh (scan) to bring up information stored in Assetwise about the piece of equipment.

Miniaturized imaging scanner captures CES Innovation Award

The Leica BLK360 has won a CES 2017 Innovation Award.  The BLK360 is a miniaturized 3D imaging laser scanner that weighs about a kilogram. It captures full-color panoramic images overlaid on a millimeter accuracy point cloud.  It shoots 360-degree high dynamic range (HDR) spherical imagery and combines it with a millimeter accuracy 3D laser point-cloud.  It is designed to simplify the reality capture process - the BLK360 has only one button. A complete 360° laser scan with panoramic image capture and transfer to the iPad Pro requires 3 minutes. BLK360 was created in tandem with a new version of the Autodesk ReCap 360 software that runs on the Apple iPad Pro.  It is intended to enable users to easily create, share and manipulate accurate 3-D models.

It collects 360,000 points per second and has a range of 60 meters with 4 mm accuracy.  In addition to a LiDAR sensor, the BLK360 includes infrared sensors for thermal imaging and 360° cameras.  It is completely wireless and streams image and point cloud data to an iPad.  With an anticipated retail price of the scanner and software is $15,990/€15,000, this imaging scanner lowers the entry barrier to precision laser scanning.

Completely automated classification of 3D scans by next year ?

Today's 3D scanners are able to generate 3D models in which humans can easily recognize objects such as vegetation, ground, buildings, power transformers, utility poles and cables, and windows and doors.  Automating this process of classifying point clouds or meshes, also known as feature extraction, will enable automation of many currently time consuming workflows in many industries.  In a panel discusion at the SPAR3D 2017 conference this year Greg Bentley, CEO of Bentley Systems, alluded to a Bentley-Microsoft project to apply machine learning to classifying reality meshes and hinted that we might see fully classified reality meshes demonstrated at next year's joint SPAR3D + AEC ST conference in Anaheim.

One example where automating feature extraction would provide big benefits is automating vegetation management for transmission lines.  Periodic transmission line inspection is mandated for North American transmission lines by NERC.  It is currently carried out by helicopter-born LiDAR which is expensive and for that reason not done more often than is required by NERC.  Full automation, which is possible with beyond-visual-line-of-site (BVLOS) autonomous UAVs and automated feature extraction of lines, pylons, vegetation and ground, would dramatically reduce the cost of inspection and enable more frequent flyovers.  The major benefits would be lower costs and a reduction in the number and length of outages resulting from vegetation transmission line interactions.

The latest 3D scanners, which are capable of creating high precision photorealistic models of just about anything, from digital terrain models and cities to detailed models of cathedrals, are getting much more affordable and accessible.  The current lowest price point is about $400 to $450 for adding 3D scanning capability to a tablet.  In the future 3D scanning and imaging may simply be built into smart phones.  Automating feature extraction from the exponentially increasing volume of 3D data, whether point clouds or meshes, is going to grow in importance.

But automating feature recognition and extraction from a point cloud or mesh remains a challenge.  Researchers have attempted to use machines to recognize features in imagery ever since computers were able to manipulate and analyze raster data.  There are cases where imagery contains a restricted set of types of objects in a known environment which makes feature recognition simpler.  For example, identifying icebergs in satellite or overflight imagery of the North Atlantic.  As another example, in a refinery or a chemical plant most of the objects are pipes, structural steel, and ducts.  For this type or environment with well-defined geometrical objects, software such as Trimble RealWorks (TRW) can identify walls, floors and ceilings inside buildings and ground, buildings, vegetation, and curbs and gutters outside.  At SPAR3D it was announced that the latest version of TRW 10.3 can even resolve power lines, which is critical for automating vegetation management for transmission lines.

I have blogged about Indoor Reality who have developed technology that can track accurately the location and orientation (six degrees of freedom) of a backpack containing LIDAR and other sensors as the operator walks through a building going into rooms, halls, up and down stairs and into other human accessible spaces.  Software tools are provided to generate intelligent triangulated mesh surfaces.  The package includes analytical software tools for automatically generating floor plans, distinguishing rooms, identifying staircases, and identifying doors and windows, in fact, everything needed for a Revit model of the interior of the building.

But for general imagery automating feature extraction typically follows the Pareto principle - extracting 80% of the features is easy, but the remaining 20% is difficult and requires most of the effort.  Currently companies such as Clearedge offer specialized software, Edgewise, for classification of 3D imagery.  Trimble has licensed Edgewise from Clearedge for construction applications and it can be used to extract pipes, conduits, structural steel, ducts, structural concrete, walls, windows and doors from 3D scans.  But the current algorithms are not able to classify everything and the process remains semi-automated - Clearedge claims that in general Edgewise can speed up the process of feature extraction by 70 %.  I am looking forward with extreme interest to seeing fully classified reality meshes demonstrated, perhaps as early as next year's joint SPAR3D + AEC ST conference in Anaheim.