Between the Poles

Construction industry poised for rapid digitization

The CG/LA Infrastructure Leadership Forum in Montreal has provided a fascinating glimpse into what is happening around the world in infrastructure from a technology, financing, and business model perspective. . From 50,000 feet the big item is that the construction industry may be at the tipping point with respect to digitization.

PPP

Probably the highlight of the event is the success of PPP projects, especially in Canada.  About one quarter of major infrastructure projects in Canada are PPP.  In the U.S. the PPP model works well in some jurisdictions, for example, the Port Authority of New New York and New Jersey.  In some states PPP is illegal and even at the federal taxation level until recently there were tax regulations that hindered the adoption of the PPP model.  The presentations made clear that there are multiple PPP models.   Several projects in Latin America were looking at PPP alternatives that were better suited to their jurisdictions.  A conference attendee from Nigeria was clearly very interested in which of the PPP models were appropriate in his country.

Investment in infrastructure

Big numbers were on virtually every speaker's lips.  The Canadian Federal government intends to invest $180 billion over ten years in infrastructure, the Quebec government $87 billion, President Trump threw out $1 trillion over five years in a state of the union speech a few days ago, Boston Consulting Group has estimated that there is a spending gap of $1 trillion annually, and the ASCE estmates that US infrastructure requires an investment of $4 million just to make up for the lack of maintenance over the past decades. Some speakers made a convincing case that these estimates are overblown. Whatever the numbers there was a consensus that there is a very large need for investment in infrastructure. 

But infrastructure, while among the top ten priorities for Americans, is not up there with jobs and the economy.  The Boston Consulting Group estimates that there are 15.4 million jobs in the infrastructure industry.  On average these are better paying than the average blue collar job.  The BCG modeling suggests that $ 1 trillion invested in infrastructure could get this up to 19 to 20 million jobs.  To get attention infrastructure has to be about jobs.

The good news is that there was a consensus that there is a lot of money looking for infrastructure projects. One speaker estimated ten times as much money as projects to invest in.  The holdups appear to be:
1) Permitting - Few are willing to commit money to projects that don't have the requisite permitting in place.  In the U.S. that process requires on average 9.5 years, but can take up to 15 years.
2) Social licence - These days it is not enough to simply get the state, federal, presidential, US Army Corps of Engineers and other regulatory permits.  It is necessary to get a social licence which means actively seeking buyin from the community.  For example, the Plains and Eastern Tennessee transmission line project has been supported by the Sierra Club and other environmental organizations as well as local landowners and other stakeholders.  Sometimes opposition comes from unexpected sources.  Clean Line's several big transmission projects have sometimes been filed against by the local utilities companies whose service territory the line crosses.
3) Business case - Investors are looking for a low risk investment with a good return.  That means that benefits to various stakeholders have to be identified and stakeholders brought into the project to spread the risk and reduce it for any one participant.

A lot of this money is private from domestic and international sources.  Japan and Saudi Arabia were both mentioned as sources of money for infrastructure investment.  Canadian pension funds are active in investing in infrastructure internationally.  To attract private funding infrastructure projects will have to generate returns, and that will drive investment in technology to improve productivity.

U.S. infrastructure market

Another highlight of the conference was Dan Rule's presentation on the Trump Transition Team's recommendation for infrastructure development in the next four years, which I blogged about previously.

Interestingly what Trump decides to do with respect to infrastructure may not be all that important. In the U.S. the federal government is not by any measure the biggest player in the infrastructure market - it only represents about 10% of the market. The big player is local governments. Local government has enormous power to choose projects, raise money and borrow money. This local autonomy is baked into the Constitution. Local governments decide on infrastructure projects and then float (tax deductible) bond issues to finance them.  The muni bond market is $3.2 trillion and is the second largest debt market.  There is good news here as well. The majority of the bond issues at the local level to fund infrastructure projects are successful in raising the necessary money through rates.State governments are also big players in the infrastructure market.  One speaker recommended to vendors and consultants looking for opportunities in the U.S. to look for municipalities who had recently successfully floated bond issues or states that had increased the gas tax or counties that has implemented a sales tax. In some states the state government and local governments are collaborating to fix all bridges in need of rehabilitation rather than inefficient one-offs by individual municipalities and counties. 

Large infrastructure projects

There are some very innovative large infrastructure projects around the world that were profiled at the conference.  I was very impressed that the long-haul transmission business appears to be coming to life in the U.S.  This business is not being driven by the incumbent power utilities, but by new players like Clean Line.  Geospatial technology underpins transmission line siting, transportation, and increasingly horizontal construction as drones, scanners, 3D visualization and other digital technology transforms the construction industry.

Technology

From a technology perspective just about every speaker referred to the lack of progress on productivity in the construction industry. Engineering and construction are second from the bottom on the IT adoption chart. But Wipro, Booz Allen Hamilton, Aconex,  Jacobs, and other vendors and analyst McKinsey believe that the construction industry is poised for the digitization revolution.  Vendors are investing heavily in technology including artificial intelligence, digital modeling such as BIM and full lifecycle BIM or BIM++ as one speaker referred to it, machine learning, drones, neural networks, virtual reality, big data, analytics, automation and hyperautomation.

GIS has been widely used by utilities for years for automated mapping/facilities management, back office records management, asset management, transmission line siting, and more recently for design and construction, energy conservation, vegetation management, mobile workforce management (MWFM), and outage management (OMS). Now, utilities are integrating GIS with automated meter infrastructure (AMI) and supervisory control and data acquisition (SCADA) systems. Intelligent design has crossed over from the office to the field in utilities, also enabled by the capabilities of GIS. Geospatial-related analytics (spatial analytics) is seen as one of the key aspects of success for electric utility operations in the smart grid era. Looking for patterns and correlations between different land, weather, terrain, assets, and other types of geodata will be increasingly important for utilities. Power-related analytics with geospatial components include network fault tracing, load flow analysis, Volt/VAR analysis, real-time disaster situational awareness, condition-based maintenance, and vegetation management.

As an example, PrecisionHawk is not only selling drone services but the geoanalytics to extract useful, actionable information. One of their targets is the construction industry and they intend to fit their solutions into existing workflows.  For example, a drone can collect aerial imagery of a stockpile, and the software provided by PrecisionHawk enables the volume of the stockpile to be calculated from the imagery without the user having to know much about drones, imagery, meshes or algorithms for calculating stockpile volumes.  PrecisionHawk also have an FAA permit for beyond-line-of-sight (BVLOS) drone flights. (I already blogged about the pilot flights that Xcel Energy also has an FAA permit for.) This suggests the next step will be drone data collection and geospatial analytical applications based on this data for linear construction projects.

Development of an open spatially-enabled BIM query language

Introduction

Building Information Models are comprehensive digital representations of buildings which include 3D geometry and associated data. They are widely used for designing, engineering, and constructing buildings (vertical BIM) and infrastructure (horizontal BIM). Owners are beginning to see value in BIM models and data for operating buildings. A BIM model includes not just the 3D geometry of all building elements and the enclosed spaces, but also semantic information representing both the type of the elements, the relationships between them (defined by the data model) and data associated with the geometric elements.  With BIM models being mandated in countries like the UK, Finland and Singapore, and in France and Russia in the near future,  and the increasing adoption of BIM by the private sector in the U.S. and other countries, the volume of BIM data is growing rapidly.  By merging models from different disciplines into common virtual models, the amount of information stored in such multidomain repositories quickly becomes large and complex even for average projects.  Centralized or distributed repositories enable the creation, integration and maintenance of large quantities of data from various stakeholders involved in the planning and building process and the wide range of specialized software tools used by these stakeholders.  Many ways of creating, manipulating and visualizing BIM and CAD models have been developed including CAD, Solibri, Navisworks, GTPPM, mvdXML, PQML, EQL, SQL, XSLT/XQuery, SPARQL, Gremlin, Linq, SDAI, and generic programming languages.  However, the development of a general, intuitive query language for BIM models including geometric and related information remains a challenge.

The information needs of individual stakeholders in different parts of the construction process are very different. For example, a structural engineer does not require all the information available in the overall building model such as the detailed specifications of a suspended ceiling. This makes the extraction of partial model subsets or domain specific views on large models necessary.  In addition regulations, energy performance modeling, and building maintenance require finding specific facilities among the building assets. 

In the last few years research has focussed on the development of an open, domain-specific query language for BIM.   Most of the this research has focussed on querying, updating and deleting data stored in Industry Foundation Classes (IFC) models.  The Industry Foundation Classes (IFC) is a vendor neutral and open model that defines more than 650 entities and a few thousand attributes.  Most industry software BIM tools  are able to export IFC models, though there is typically some information loss. The bimserver.org platform is an open repository of IFC models and already provides a Java API to extract partial building information models from a repository.

Open query languages

Many existing database applications use the Structured Query Language (SQL), an ANSI/ISO standard with spatial extensions standardized by the Open Geospatial Consortium (OGC), as the standard language. SQL makes it possible to create, read, update and delete records in a Relational Database Management System (RDBMS).  But the applicability of SQL in a BIM context is limited.  SQL is based on operations on data stored in tables.  Application to IFC, as a BIM example, would require storing 600 entities along with their thousands of attributes in individual tables.  Severe performance and scalability issues are associated with using SQL for BIM.  In addition the complexity of IFC-based queries would rapidly become too complex.

The W3C Data Activity has defined a collection of technologies comprising a “web of data” where information is easily shared and reused across applications.  Key parts of this technology stack are the RDF (Resource Description Framework) data model, the OWL Web Ontology Language and the SPARQL protocol and RDF query language which has been standardized by the W3C.  The Resource Description Framework (RDF) is a directed, labeled graph data format for representing information on the Web.  RDF and SPARQL  enables interoperability among distributed heterogeneous information systems making possible queries that span several repositories and link information from different sources.  Specifically in a BIM context it enables ad hoc queries of sub models that reside in repositories maintained by the different stakeholders in a building project.

Spatial queries

BIM data models can be simple or complex.  A complex data model includes logical relationships between building elements, such as windows and walls, walls and storeys, and so on.  These relationships are created at design time.  A simple model may not include any explicit relationships. For example, there may not be any way to determine which windows are associated with a wall except visually.  Spatial queries enables relationships to be determined after the creation of the BIM model.  For example, a spatial query can find all the windows in a wall, all the walls of a storey, or all the HVAC and plumbing facilities in a storey.

The efficient implementation of spatial BIM queries requires 3D spatial indexing such as octree and multi-dimensional r-tree indexes and algorithms.  It has been known for a long time that spatial indexing is a key technology enabling efficient GIS systems.  Very early on in the development of spatial indexes it was realized that these are domain specific.  Several GIS systems support an extended spatial SQL based on quadtree or r-tree indexes.  These systems evolved historically from a 2D geometric representation approach.  However, the functionality for processing 3D geometry in many GIS systems is restricted and the performance of the available tools is unsuitable for complex building data sets.  Therefore, it has been argued that a specialized tool is required to handle queries against elaborated BIM models containing vast numbers of elements and detailed geometry representations.  Such a query language enables the spatial analysis of BIM models and the extraction of partial models that fulfill certain spatial constraints. Among other features, it includes operators that reflect the topological relationships between 3D spatial objects including "within", "contain", "touch", "overlap", "disjoint" and "equal" in 3D space.  Both octrees and multi-dimensional r-tree indexes have been investigated as a basis for an efficient BIM query language.

Open BIM query language

A BIM query language, dubbed BIMQL, has been proposed as an extendable, open, domain specific query language for BIM models. It can be used to select and update partial aspects in building information models.  A limited implementation of the language has been built on top of the bimserver.org platform.  It is not complete, in particular it does not support spatial queries, but has provided a useful vantage point for future research, discussion and development on the development for end-user interfaces to complex BIM models as well as the composition of service-oriented architectures.

Unlike traditional data modeling approaches which are limited by the scope of their underlying schemas, RDF and related technologies can provide an open and common environment for sharing, integrating and linking data from different domains and data sources.  In comparison with domain-specific query languages, SPARQL is especially applicable for scenarios when data from multiple sources are needed.  Some important scenarios include

• All building objects should be classified according to the local building code
• Where are the locations of companies which produce the materials used in walls placed in the hall ?

The interesting thing about these scenarios is that they require the BIM geometry and data plus other external data. This is the type of query that the W3C semantic web technologies are designed to support.

In the AEC domain, a typical use case is data integrity validation against BIM standards or building code requirements.  A Dutch example is the Dutch BIM standard requires that all building elements should be associated with appropriate building storey.  Provided that the building model is represented in the standardized ifcOWL, it is possible to check this spatial containment relationship using off-the-shelf SPARQL implementation.

Currently SPARQL does not contain the vocabularies or functions required by many use cases in the construction domain.  Recently researchers have proposed extending SPARQL with a set of domain-specific functions for the construction world. There are precedents for this strategy.  In 2011 the Open Geospatial Consortium (OGC) released GeoSPARQL, a geographic query language for RDF data, which extended SPARQL with geospatial functions that enabled spatial queries.  The advantages of this approach is that not only are the extended AEC functions compatible with existing SPARQL environments, but they can also be used to query building data combined with data from other fields.

But standard SPARQL can only crawl the data graph according to the structure of the BIM data model.  There are many use cases that depend on geometric relationships that cannot be handled by off-the-shelf SPARQL. To address this SPARQL has been extended to include spatial operators "touches", "intersects", "disjoints", and "contains" and well as other domain-specific functions.

An example of a simple spatial query using a spatial extension is finding all walls which contain or adjoin doors.  This requires the spatial function "touches" to find doors that are in or near walls.

As in the case of geospatial data, additional domain-specific functions are required to provide the necessary functionality.  For example, functions were added for different ways of measuring distance (spt: defines the spatial extensions namespace)

• spt:distance is a function to return the shortest distance between two products in 3D space.
• spt:distanceZ is a function to return the shortest vertical distance between two products.
• spt:distanceXY shortest distance between the projections of two products on XY plane

Another example is finding all windows that are close (less than 300 units) to a column.

Implementation and testing

As a proof of concept, the extensions were implemented
based on the open source Jena ARQ query engine.  An objective was to try to  minimize low level coding to make the function implementations more portable and more transparent for public reviews.

As a practical example, a case study of building code compliance checking using the International Building Code 705.8.4 (IBC 2006) was tested successfully using a BIM model of a building with 4 floors.

The researchers report that the main limitations of the open SPARQL-based approach they have proposed relate to the way it was implemented.  The implementation is not completely portable and performance could be improved, by integrating existing geometric libraries, for example.   Future research will address these issues and investigate more use cases to extend functions for specific domains and to combine more data from different sources in the construction industry.

Sources

Querying Linked Building Data Using SPARQL with Functional Extensions, 2016, Chi Zhang and Jakob Beetz

Processing of Topological BIM Queries using Boundary Representation Based Methods, Simon Daum and André Borrmann, Advanced Engineering Informatics, Volume 28, Issue 4, October 2014, Pages 272–286

BIMQL – An open query language for building information models, W. Mazairac and J. Beetz, Adv. Eng. Informat. (2013)

Topological analysis of 3D building models using a spatial query language, André Borrmann and Ernst Rank, Advanced Engineering Informatics 23(4):370-385 · October 2009

Thanks to Léon van Berlo of TNO for pointing me to this research.

Big geospatial data workshop

The rapid increase of sensing and mobile devices is an unstoppable information source that is transforming human lives in incredible ways. Every two days the human race is now generating as much data as was generated from the dawn of humanity through the year 2003. Big Data is driven by the real time collection, analyses, and visualization of vast amounts of the information most of which has a geospatial or location component. The data was measured somewhere or it describes events at some location. Processing this enormous and varied stream of geospatial information with new big data techniques will bring new benefits to many applications in our world.

The Location Powers: Big Data workshop will bring together leading developers of big geospatial data systems. Participants will not only hear the latest developments but will contribute to the collective discussion advancing the use of big data computing techniques applied to geospatial data with results in many applications.

A public webinar will be held on Wed 3 November at 10:00EDT to share the Big Data Whitepaper and outcomes of the Big Data Summit

Tuesday 20th September

Meeting Location:
University of Central Florida Executive Development Center
36 West Pine Street
Orlando, FL, 32801 USA

#LPBigData

Future of Utilities: Utility asset management will be based on shared geospatial platform

Currently geospatial data and technology are used tactically by electric power utilities for many purposes including outage management, vegetation management, disaster management, renewable energy facility siting, universal electrification planning, asset management, and energy density mapping, to name just a few.  But the smart grid is fundamentally transforming the role of geospatial data and technology in electric power utilities. According to a report from Navigant Research “The smart grid is all about situation awareness and effective anticipation of and response to events that might disrupt the performance of the power grid. Since spatial data underlies everything an electric utility does, GIS is the only foundational view that can potentially link every operational activity of an electric utility including design and construction, asset management, workforce management, and outage management as well as supervisory control and data acquisition (SCADA), distribution management systems (DMSs), renewables, and strategy planning.”  In other words the smart grid makes geospatial strategic.  Location is the basis on which data is organized and spatial analytics is how information is extracted from the data.  As Bradley Williams has so succinctly put it, "spatial analytics is becoming a key technology for electric utilities because everything a utility does - customers, assets, and operations - involves location."

At the Future of Canada's Utilities Summit in Toronto I had the opportunity to chat with Alexander Bakulev, VP of Strategy and Assets at METSCO, in Mississisaug, Ontario about asset management, location and GIS, and the future of utility IT.  Alex has deep experience with asset management at utilities in Ontario and as VP of Strategy he brings a forward-looking perspective to the changing role of location and GIS in asset management in electric power utilities. 

Alex, what type of asset management service does METSCO provide to your customers ?

We provide a framework for asset management based on a methodology, risk-based asset management. When a decision needs to be made such as whether to replace an asset or maintain it and run it to failure, those decisions are made by engineers to mitigate the risk of failing as well as for safety and environmental reasons. We enable our customers to translate those risks into dollars so they can compare dollars to dollars and make an economic decision whether to replace or maintain an asset.  Of course, this has to be done within a regulatory environment.  In Ontario this means PSAB 3150 and ISO 55000.

Can you give a perspective on how advanced your customers are with respect to asset management ?

Asset data

We have a maturity hierarchy that helps us understand where a utility company is in managing its assets.  When we begin working with a customer, the first thing we look at is their asset data, what do they have, in what media, and where it resides.  Companies have different maturity in term of the asset records they have and how they manage them. Usually what we see is that companies may have good inspection data, but the information is stored in a decentralized mode -  paper documents, Excel spreadsheets, special lab-testing software files, and so on - but not in a centralized system. In many cases we have to compile paper documents and files to order to combine it all into a centralized asset management system.  But we see that is changing. Companies are starting to invest in document management systems and capturing data digitally. They are also collecting data with mobile devices in a way that goes directly to the online ERP system.  In addition newer equipment is smarter and able to report inline so it's a paperless process.

Risk of failure

Once you have the basic information about the condition of each asset, the next step is to try to predict each asset's failure probability. We want to understand what's the chance that they will fail next year, after 5 years,  or 10 years.  If the company has historical data on failures including the type of asset, when it failed and why it failed, we can use that information to statistically predict failures.  To do this we first find data about which assets have been replaced. Then we try to determine the reasons why the assets were replaced. Often this information is even less available than condition information. With this information we can estimate the expected failure rate for existing assets.

Consequence of failure

Once you know the risk for every asset to fail, the next step is to look at the consequence of failure.  We not only look at what is being fed downstream from an asset such as a transformer, but also at the protection schema on the feeder.  For example, when a radial feeder fails, the result is a minimum of 4 hours outage.  If it the feeder is a loop design, the outage may only be 30 minutes. If the feeder is automated, the outage could be seconds.  And it may not be just power that customers lose, there could also be environmental and safety issues. For example, if a corner pole fails, the consequences are more serious than a pole in a street line. There are also financial factors. How fast do you need to replace the asset ? Do you keep it in stock or order it ?  For expensive assets, it is very expensive to keep spares in a warehouse.

Monetized risk

Once we know the condition of each asset, the probability that it will fail, and the consequence of failure,  we can monetize the risk. With monetized risk, when we have to make a decision, we make an economic decision.  Maybe it's better to run to failure, or increase the frequency of maintenance to watch the asset closer and accept the risk of failure for the next 5 years and after that replace it, or if it is high risk, replace the asset immediately. 

It sounds like you can classify utility companies based on the data they have and how they use it to make decisions ?

We have a maturity model that helps us determine where a customer is so we can determine how best to help them.

Run to failure -  Assets are only replaced when they fail.  From the perspective of financial costs, this alternative looks good, but if you look at customer value, the cost is very high.  There are many unplanned outages of relatively long duration.

Age-based - In this case assets are replaced based on the age of each asset.  In this case investment requirements can be three times higher than run to failure, but the customer is benefiting by improved service.  There are much fewer unplanned outages.

Condition-based - With this approach you only replace assets that are in poor condition. Investment requirements are less than age-based, but still quite high compared to run to failure.  Customer value is similar to age-based.

Risk-based -  For example, you may decide to run 70 % of assets to failure in those areas where there is low risk, but for the 30% of assets that are high risk, you try to avoid any failures. This requires focussed investments.  Investment requirements are significantly lower than condition-based, but still higher than run to failure.  But with this alternative, customer value is optimal within the budgeted investment level.

Where are most of your customers also in the maturity model when you first start working with them ?

When we first begin to work with them, most are somewhere between age-based and condition-based. Many are making age-based decisions, but are trying to move to condition-based asset management for many asset classes.  Surprisingly there are some large companies that still follow a run to failure approach.  We have customers that have gotten to the highest level, risk-based asset management, and we have helped them get there.

What size utility finds it easiest to do this ?

In our experience, mid-sized utilities - with half a million to a million meters - find this transition easiest.  They are more flexible.  For large utilities the transition  is challenging because it they have to change what they have been doing for the past ten years, mostly age and condition-based asset management, and they have implemented large expensive systems to help them do that.

How does location and GIS fit onto this ?

Location and GIS are fundamental in several areas of asset management.  First of all geospatial technology provides a communication tool.  Most critically geospatial visualization is used to justify investments. It makes it possible to show stakeholders high risk areas. It makes it possible to show people recommended investments and where to invest.  

Secondly, it helps to plan for the future.  By visualizing changing demographics it helps identify what parts of a city require investment. For example, if in the next 10 years most of the work is going to be done in a particular rapidly growing area of the city, it can be planned for by creating a yard in the area, for example.

We also use GIS for core planning.  When engineers are designing the electric power network for a new subdivision, they require very detailed information about existing electric power assets in and near the area.  This information is collected from many different sources and visualized on a map.  The engineers can look at the risk profile of every existing asset on the map and start planning which assets need to be replaced immediately, which in five or ten years, and where they need to install new equipment.

Another critical area for GIS is connectivity,  Connectivity is about which customers are connected to which service point, which service points are downstream from each transformer, and which transformers are on each feeder.  Connectivity information is essential because it enables utilities to know who is affected by a planned or unplanned outage. It also provides a link between customers and infrastructure for financial planning purposes.   Many utilities don't have reliable connectivity information or they think they do but really don't.  GIS is the primary tool that we use to establish or correct connectivity.

Another critical area that is growing in importance for GIS is spatial analytics.   GIS can help estimate the condition of assets and identify high risk assets. For a simple example, in a tabular ERP system, all poles are the same, but in the GIS we can see some poles are corner poles which are higher risk.  As another example, if an asset is near a highway which is salted in winter, this can affect the condition of the asset. Here we are getting into predictive analytics where we can make estimates based on proximity to things that may affect the expected lifetime of assets.  In this case assets near roads with salt in winter are at higher risk.

What's your perspective on the future for utility asset management  ?

Many companies are beginning to see the advantage of the mapping environment for asset management. In the past asset management was pretty much done with Excel or similar table-based tools. What we are seeing now is that increasingly decisions are being made based on maps, increasingly interactive maps.  There are several drivers for this.

The first driver is simplification.  Asset management is simpler when you put everything on a map.  And for the new generation of engineers, maps facilitate knowledge transfer - it 's easier for them to design, manage and maintain the grid with assets visulaized on a map.

The other driver is grid complexity.  With the smart grid and distributed energy generation, the grid is getting much more complicated. There is a lot more data that needs to be analyzed.  Thefe are also increasingly complex reporting requirements from the regulator.  There should be one environment that can handle all of this - data management, analytics, and regulatory reporting.  Maps help alleviate complexity and achieve simplification.

Do you foresee in the future that utility applications such as outage management, customer information system, crew management and dispatch systems, and others will be best of breed or an integrated solution from a single vendor ?

The future as I see it is all about map-based decisions - where are the outages, where are our marketing opportunities, where are assets likely to fail, where do our employees live, where are the customers that have lost power, where should we position crews in preparation for a storm ?   To enable this to happen, I believe the future for utility IT is a map-based platform which applications like asset management, design, outage management, and disaster management from different vendors can plug into and share information in a shared geospatial environment.

HxGN Live: Hyper-spectral camera mounted on a UAV for precision agriculture and other applications

The typical multi-spectral camera used on satellites and airborne platforms captures three or four bands - typically the Landsat bands: green, red and near-infrared.   But hyperspectral cameras can capture 270 bands and are light enough to be mounted on a UAV.  At the annual Hexagon conference HxGN Live in Anaheim, Headwall Photonics gave an overview of the capabilities and advantages of hyper-spectral remote sensing.  Headwall and Leica Geosystems have partnered to offer a complete UAV-based hyper-spectral package which includes the Aibotix X6, Headwall hyper-spectral camera, storage, IMU/GPS, and software. 

A hyper-spectral camera passes the light it captures through a diffraction prism that breaks the light into a spectrum that ranges from infrared through visible to ultraviolet.  The Headwall camera records 270 bands ranging from the infrared through the visible spectrum. This broad spectrum allows the chemical composition of materials to be detected.  Spectral fingerprints can be defined based on chemical composition which enables detection of different species of trees, stages of crop development, disease incidence in crops, specific chemical compounds in mining tailings, and many other materials.  The Headwall camera can capture up to 300 frames per sec.  That generates a lot of data.  The key to extracting information from this volume of imagery is spatial analytics software which is part of the package.

The Headwall Leica Geosystems package is an integrated multi-rotor UAV and visible and near infrared (VNIR) hyper-spectral imaging system with software control.  It includes Headwall’s Nano-Hyperspec VNIR sensor with the Leica-Geosystems Aibot X6 multi-rotor UAV. The Nano-Hyperspec VNIR Sensor captures bands in the wave length range 400-1000 nanometers (nm) with a 12 bit pixel and a maximum frame rate of 300 Hz.  It also includes 480GB of on-board storage and an attached GPS/IMU. 480 GB corresponds to about 130 minutes of imagery at 100 fps.

Typical applications of hyper-spectral imagery are
precision agriculture, mineral exploration, and civil engineering.  Specific examples of the application of the technology include marijuana plant detection, crop health monitoring, citrus species mapping, invasive species monitoring, growth cycle and monitoring of crop yield, mine tailings monitoring, storage tank leakage detection, environmental monitoring, disaster monitoring, and mapping urban areas. 

A fascinating example is monitoring crop health where a scan of a field shows areas where the plants are not doing well.  This enables the selective application of fertilizer to only those areas where it is required.  Similarly a hyper-spectral scan reveals areas of insect infestation, which enables the selective application of insecticide.  Another application that would interest grape growers in the wine industry is using hyper-spectral imaging to identify and monitor leaf roll virus infestations.

Hyper-spectral imaging can also be used in an urban environment roof type mapping, insurance assessments, identifying types of road and monitoring road conditions, monitoring tree conditions in urban forests, and for urbanflash flood and disaster preparation and monitoring.

The key differentiator of hysper-spectral imaging is much greater precision in recognizing specific chemical compounds whether in plants, rocks, mine tailings, or urban environments.  However, the price is managing a much greater volume of data. Spatial analytics is key to extracting information from the huge volume of data collected in the field.

Citizen geodata science an emerging big data trend

Gartner's latest hype cycle identifies a new trend "citizen data science" that has important implications for the geospatial sector.

At the recent Canadian Institute of Geomatics (CIG) Ottawa Branch Workshop on Big Data: What is it? Technology and Applications, Monica Wachowicz from the University of New Brunswick pointed out a brand new entry Data Science in the Gartner Hype Cycle 2014 that wasn't in the Gartner Hype Cycle 2013. Then in the Gartner Hype Cycle 2015 Data Science and Big Data have disappeared and a new entry with the intriguing name Citizen Data Science has appeared.

2013 Big Data

2014 Big Data and Data Science

2015 Citizen Data Science

What is data science ?

Monica referenced research by Harlan Harris et al “Analyzing the Analyzers” An Introspective Survey of Data Scientists and Their Work which surveyed data scientists to try to determine what data scientist do.  They divided data scientists into five categories based on their self-ranked skill sets; Statistics (which includes spatial statistics), Math/Operations Research, Business, Programming, and Machine Learning/Big Data, and four categories based on their respondents' self-identification; Data Researchers, Data Business people, Data Engineers, and Data Creatives.  

• Data Business people are most focused on the organization and how data projects yield profit.
• Data Creatives excel at applying a wide range of tools and technologies to a problem, or creating innovative prototypes at hackathons.
• Data Developers are focused on the technical problem of managing data — how to get it, store it, and learn from it.
• Data Researchers backgrounds are in statistics or mathematics, 75% of them have published in peer-reviewed journals, and over half have a PhD.

The resulting landscape of data scientists shows a broad range of skills distributed among the self-identified professional categories.  The researchers found that

• a defining feature of data scientists is the breadth of their skills, and their ability to single-handedly do at least prototype-level versions of all the steps needed to derive new insights or build data products.
• the evidence points in favor of a scientific versus a tools-based education for data scientists. 
• 70% of the respondents had at least a Master’s degree. 
• 40% had undergraduate degrees in scientific fields, specifically, physical or social sciences (but not mathematics, computer science, statistics, or engineering).
  

The survey showed that the most successful data scientists are those with substantial, deep expertise in at least one aspect of data science; statistics, big data, or business communication.  The authors concluded that data science is a collaborative and creative field, where the successful professional is able to work with database administrators, business people, and others to get data projects completed in innovative ways. 

What is citizen data science ?

One of the biggest drivers for a new type of data specialist is that there are simply not enough data scientists to satisfy the demand from virtually all sectors of the economy.  Gartner has recommended cultivating “citizen data scientists” which they identify as people on the business side who may have some undergraduate mathematics or social science background and who can be assigned to exploring and analyzing data with the appropriate software tools.  Last year Gartner predicted that the demand for citizen data scientists will increase five times more rapidly than that of the highly skilled data scientists Harlan Harris studied.

Software tools are critical for helping citizen data scientists find real insights and avoid simple statistical mistakes.  For example, a technology that Gartner identified is "smart data discovery" which is a next-generation data discovery capability that provides business users or citizen data scientists with insights from advanced analytics and helps them avoid some of the common statistical pitfalls.

Citizen geodata science

There is at least ancedotal evidence that there has been for some time a trend toward "citizen geodata science".  At the India Geospatial Forum in Hyderabad in 2014, I moderated a session on electric power.   One of the speakers was Arup Ghosh, Chief Technology Officer at Tata Power Delhi Distribution Ltd (TPDDL) who presented his perspective on implementing geospatial technology in a private utility.   The geospatial group at TPDDL had about 60 field personnel and 18 analysts and support staff,  none of which had an educational background in geospatial data and technology.  Twelve were electrical engineers and the rest were people with electric power experience.  All learned geospatial data management and simple analytics "on the fly".

At a recent GoGeomatics Social Jonathan Murphy gave a presentation about his experience working in Northern Alberta preparing terrain for seismic surveys.  Geospatial data and technology was used in all aspects of field operations.   The staff were experienced in seismic surveying, winter drilling programs, wildfire management, and road and facility construction, but had minimal education in geospatial data management and analytics.  They also had picked up enough geospatial knowledge "on the fly" to do their jobs.

These are two examples of what could be called "citizen geodata science" which I suspect is part of Gartner's citizen data science trend and is almost certainly growing more rapidly than traditional "geospatial data science" as geospatial technology is adopted into vertical industries.  The challenge is how to reach these people, perhaps through MOOCs, community colleges, conferences focused on vertical industries and including geospatial technology vendors, presentations and hand-on training, or vendor marketing to help them avoid the common pitfalls of geodata management and geoanalytics.

Automating the process of comparing digital models to reality

At the Spar 3D Expo and Conference in Houston I had the opportunity to chat with Nicolas Arnold, VP of Product Development at SKUR.  SKUR is a young software company that has just released its first product, which automates comparing a 3D digital model to reality as represented by a point cloud from terrestrial scanning technologies (typically LiDAR).  The scan can be of an existing structure or a new structure in construction. The variances are prioritized and can be inspected in an industry standard 3D visualization application.

Nicolas, SKUR provides technology that finds the differences between digital models and point clouds from scans.  What are some of the applications of this technology ?

We are in the business of comparing 3D building models with point clouds that are registered to the building model. We do this for two very specific use cases.

The first use case is what I refer to as construction QC.  Let's say you have a design model and during construction you capture a point cloud by laser scanning or some other technology.  What you are interested in is comparing the design model to what was actually built out in the field. To do this you simply upload the design file and the point cloud, specify three tolerances that are used to prioritize the differences between the model and point cloud, and then run our cloud-based application.  The application outputs colour-coded deviations, which you can visualize in Navisworks.

The second use case is what I refer to as model QC where you have the reverse situation.  Let's say there is an existing structure, a building or almost anything.  But there is no digital model and you need a model for a renovation or retrofit.  You can use a scan-to-BIM process to scan the building and generate a point cloud from which a model can be created.  The model creation could be done by your own people, by a third party provider, or by an automated process such as Clearedge.  In this use case what you are interested in is how  well the model fits with reality. You follow the same process, uploading the model and the point cloud, to compare the model to the point cloud to see where there might be modeling errors and other types of deviations.

These are the two use cases that SKUR serves. We target the sophisticated construction contractor (GC).  One application of the technology for a GC is to perform construction QC on their projects to look for problems.  Another application is pre-construction, when the GC is considering bidding on a renovation or retrofit project and wants to provide a digital model to the architect.  The GC wants to be confident that the model represents the structure accurately so he can tell the architect that the model is a reliable representation of reality. These are the folks that we are targeting for these two use cases.

Does SKUR provide data capture services ?

No, we are a software company.  If you are interested in this technology, you can create an account, it's up to you to capture the point cloud data however you want to do it, upload it onto our web-based system together with a digital model and then run our application.   You then download the variances and visualize them using an industry standard tool. We don't do any consulting.

I'm interested in how SKUR would be used during construction. What if the construction is only 10 % or 20 % complete, do I have to specify which part of the incomplete structure I have scanned ?

We have very smart algorithms that figure out which part of the structure to look at.  For example, if you bring us a point cloud of just the corner of a building and the model of the entire structure, we only focus on that corner of the building.  Another example is a building that is 12 stories high, but I only have a scan of the basement. Our software is smart enough to figure out the part of the model that relates to the scan.

What about the opposite case where during a scan-to-BIM process for an existing building, I have not finished the model so I only have a model for part of the building ?

It's basically the same algorithm. The software is smart enough to figure out which part of the scan relates to the partial model.  When you run our software, you simply upload the model and the point cloud and specify the tolerances.  The process is the same for construction QC or for model QC. From the perspective of the software, it's really the same thing.

Where does your software run, on a desktop or in the cloud ?

We are completely web-based, that is to say, cloud-based. Our software require heavy-duty processing that would not have been possible 5 years ago. It is all in the cloud and it's high performance multi-core computing.   We bring in the largest computers that are available on the Amazon cloud for these problems and they can generate the variances in a few hours.

What types of digital models do you support, 3D CAD, Revit, Bentley BIM, or IFC models ?

We are a young company, and we've just started.  Our first release supports 3D DWG models. But I think Revit models are the next logical step. And then we also plan to support the Bentley world.  The 3D DWG model really can be anything, with any level of complexity. I have had people asking me about things other than construction, like planes, such as an Airbus. I talked to a gentleman yesterday, who wants to use our software for a golf course.  As long as you have a DWG model and a point cloud that's registered to the DWG, we can find the variances between the two.  The registration between the two is critical. It has to be done accurately.

How difficult is it to choose the tolerances ?

It's very easy.  You choose three values corresponding to different degrees of variance which are colour coded, green, yellow, and red in the output.

How do you visualize the variances, I imagine for a large project there could be thousands of these ?

SKUR output is a visual representation and report of the detected locations where there are discrepancies showing the precise variance from design and, very importantly, the direction of the variance.  We colour code the discrepancies so you can spot the problem areas very rapidly. For example, in the graphic red and yellow coloring indicates degree of variance. Arrows show the direction of issue. In our experience most of the variances are green, so you can filter these out and just look at the ones colour-coded red, which are the big discrepancies.  This allows you to focus just on those important areas.  We're like an insurance company.  Most of the time it's fine, but when it's not you need to know.

Where did the algorithms come from  and where was the implementation done ?

We've invented everything from scratch in Oakland, California

I can imagine this technology being of tremendous value in many different sectors of the construction industry as well as in non-construction sectors.  Which sectors and geographies are you targeting as the market for your technology ?

The sector we have targeted so far are construction contractors (GCs). But going toward we will be increasingly focussing on owners.  We think that our software provides a view into the construction process that enables them to see during construction how well the designs by the architects and engineers compare to what is actually being built.

We are focussing on North America, but we also have clients in Europe. Europe is definitely high on our list, but as a young company we are focusing on close to home first.

If people want to try your software, how do they engage with you?

It's very simple. They sign up and create an account.  Then they can upload a model and a point clouds and hit the magic button. We do handholding in the beginning to help new customers.  But we don't do any data capture and or any data preparation. We simply provide a web-based software service.

Study suggests U.S. responsible for major part of atmospheric methane rise over past decade

 Methane is one of the more potent greenhouse gases for global warming.  The concentration of methane in the atmosphere stabilized from about 1999 to 2007, but since 2007 began rising again.  A recent study suggests that the more than 30% increase in U.S. methane emissions over the 2002–2014 period could account for 30–60% of the global growth of atmospheric methane seen in the past decade.

When Prime Minister Justin Trudeau visited President Barack Obama in March, one of the topics that was discussed was joint action on climate change. Specifically, Canada and the United States committed to new actions to reduce methane pollution from the oil and natural gas sector.

In March, 2014 the U.S. Administration issued Strategy to Reduce Methane Emissions, which outlined steps to cut methane emissions.  In the summer of 2015 the U.S. Environmental Protection Agency (EPA) proposed standards to directly address methane from new and modified sources in the oil and gas sector with the goal of reducing methane emissions from the oil and gas sector by 40 to 45 percent below 2012 levels by 2025. But studies from the Environmental Defense Fund (EDF) and others have provided significant new data on methane emitted by existing operations in the oil and gas sector that show that "methane emissions are substantially higher than we previously understood" in the words of Gina McCarthy, EPA Administrator.

In 2012, the EDF initiated a series of 16 independent projects designed to find out how much and from where methane is escaping into the atmosphere across the entire natural gas supply chain: production; gathering lines and processing facilities; long-distance pipelines, storage, and local distribution; as well as some end users using natural gas, commercial trucks and refueling stations. This investigation involved nearly 100 research and industry experts.

Scientists collected data using aircraft and ground-based platforms in the Texas Barnett Shale. Using this data they estimated regional and facility-level methane emissions and found that regional methane emissions are 50 percent higher than estimates based on the Environmental Protection Agency’s Greenhouse Gas Inventory.

Methane emissions from local gas distribution systems were also investigated it was and found that methane emissions from local natural gas distribution systems are significant.  The EDF and Google Earth Outreach released interactive maps that show methane leaking from pipelines under city streets.

The Denver-Julesburg Flyover Study found that methane emissions were three times higher than estimates derived from EPA data.

A study assessed the spatial distribution of anthropogenic methane sources in the United States by combining comprehensive atmospheric methane observations, extensive spatial datasets, and a high-resolution atmospheric transport model. Based on the results of this analysis the authors concluded that the EPA underestimates methane emissions nationally by a factor of ∼1.5 and regionally by ∼4.9.

A recent study which analyzed GOSAT satellite and surface observations of atmospheric methane suggests that U.S. methane emissions have increased by more than 30% over the 2002–2014 period.  This contrasts with EPA estimates which indicate that emissions remained flat over this period.  The increase is largest in the central part of the country. The spatial pattern of the methane increase seen by satellite does not clearly identify the sources of the increase.  The renewed growth in atmospheric methane between 2005 and 2010 is estimated to be due to an increase in global methane emissions of 17–22 teragrams per year  (Tg/yr). The new results suggest that increasing U.S. anthropogenic methane emissions could account for up to 30–60% of this global increase.  The authors state that a better understanding of U.S. anthropogenic methane emissions, particularly those from the livestock and oil and gas sectors, is needed.  The U.S. EPA inventory gives only national totals, so identifying sectors responsible for the increase is difficult. A spatially resolved version of the U.S. EPA inventory is currently under development.

The EPA's next annual methane report is due to be published by April 15 and it will be interesting to see what the EPA is doing to address the significant difference between its estimates and the estimates from academic researchers.

The shale oil and gas sector in the U.S. has been largely unregulated.  In 2005 Congress passed the Energy Policy Act of 2005, which was an omnibus energy bill that included the so-called "Halliburton loophole" which exempted hydraulic fracturing from protections under the Clean Air Act, Clean Water Act, Safe Drinking Water Act, and CERCLA, which means that fracking has not been regulated at the federal level.  

Since Prime Minister Trudeau's visit to President Obama the EPA has initiated a formal process to require companies operating existing oil and gas sources to provide information to assist in the development of comprehensive regulations to reduce methane emissions.  The EPA is issuing an Information Collection Request (ICR) to gather information on existing sources of methane emissions, technologies to reduce those emissions and the costs of those technologies in the production, gathering, processing, and transmission and storage segments of the oil and gas sector.  The goal is to take a closer look at regulating existing sources of methane emissions, in other words, to set more aggressive rules for reducing methane emissions from existing oil and gas facilities.

EPA intended to begin the ICR process In April by signing a draft information collection request that will be made available for public comment. The agency will revise that draft as necessary based on comment and send it to the Office of Management and Budget for additional review and input. Once the collection request is approved– which can include surveys and required emissions monitoring – it will go to industry, which is required to respond and attest that the information it provides is accurate. EPA’s goal is to receive the first phase of information later this year. 

The EPA began the ICR process May 12, 2016.

Spar 3D: 3D mapping the inside of multi-story buildings with a backpack

At the Spar 3D Expo and Conference in Houston The Woodlands, Texas, Avideh Zakhor from the University of California Berkeley was onsite with some remarkable products from her new startup Indoor Reality.  I blogged previously about the technology when it was still under development.  The "man portable system" she has developed is a 32 pound backback with a collection of sensors.  The most important feature of the technology are sensors and algorithms for accurately tracking the six degrees of spatial freedom which determine the location and orientation of the backpack in the building.  In addition there are sensors for recording optical and thermal point clouds. She has also developed a handheld device that supplements  the capabilities of the backpack.

The technology can track accurately the location and orientation of the backpack as the operator walks through the building going into all rooms, halls, up and down stairs and into other human accessible spaces.  Software tools then generate different 3D data products including optical and thermal 3D point clouds, a colourized 3D point cloud, surface texture map,  and triangulated mesh surfaces.  The package includes analytical software tools for automatically generating floor plans, distinguishing rooms, identifying staircases, identifying windows, finding heat sources such as computers and human occupants, and identifying lights. Surface textures can be added to create a 3D photo-realtistic rendering.  The sensors and software can capture and generate all required input for the Department of Energy's EnergyPlus energy performance analysis program.  It has GPS and can generate real world coordinates.  Indoor Reality is an ESRI Emerging Business Partner.

I had a chance to sit down with Avideh to talk about her new company and the remarkable products she was demonstrating at SPAR2016.

Avideh, what are the key features of the products you are demonstrating here that distinguish them from some of the other devices vendors are showing here at SPAR ?

We have developed a unique way to track where the operator is in a building and the orientation of the backpack.  We have a lot of sensors. That is the key to all of our intellectual property (IP). It's called sensor fusion and most of our IP is in that area. It does not suffer from drift. It tracks accurately where you are in the building even if you go upstairs. You can even go up in an elevator because it has a barometer.  In addition we detect wifi.  It can tell you what the wifi strength is in different parts of the building. We have an infrared sensor and we have developed an app that will give you real-time occupancy of the building, for example, we can tell you there are four people in this room, two people in that room. We have patents on that.

Based on the data collected by the operator's walk through the building, we allow users to virtually navigate through the space and spatially collaborate with each other through our web based visualization tool. During virtual navigation, you can look around in 3D as if you were there.  That's our second data product. We don't just record dumb point clouds.  We generate intelligent meshes.  Not only can I can pan around, look up and down at each location, but the shape of the cursor shows you the orientation of the surface - for a horizontal surface it's a circle, for vertical surfaces it's a square.  The cursor gets bigger as it gets nearer, smaller as it goes farther away.  We know the normals of all the surfaces because of the 3D mesh behind what you see.  We recognize doors and windows.  Of course, we can measure distances and we allow users to annotate objects and locations.   So this mesh is very intelligent, so intelligent that we can automatically generate a Revit model with a floor plan, rooms, ceilings, doors and windows. That is our third data product.

How do you foresee people will use the tool ?

We went to the University of Illinois just a month ago and we did an energy model for a 14 story bldg. We can do 3d geometry, we can detect lights automatically, we can detect windows automatically, and we can detect AC plugs automatically.   This gets fed into an energy simulation model which is a complete model for an Energy Plus performance audit.  Typically when people do an energy audit they are only interested in the lowest hanging fruit, for example, replacing incandescent light bulbs with LEDs.   This is because collecting the data for a full Energy Plus model has been time consuming.  With the Indoor Reality backpack you can collect whatever you need for a complete Energy Plus audit very simply. For example, we can superimpose the infrared record on a ceiling to show a hot area which is very likely a hot water pipe behind the dry wall.

We have just come out with a visual building documentation tool for stakeholders to collaborate  with each other throughout the lifecycle of a building - during design, construction and operations and maintenance.  If I want to remodel a building using an architect, designer and contractor they can communicate with each other by leaving notes for each other in the virtual building.  For example, I can put a tag on a door for John the contractor. "Please fix this door."  John can come along in the virtual building and edit that note "Avideh, I can do this Tuesday."  So this makes it possible to collaborate virtually on a building project in 3D,  in much the same way that we can collaborate on a Word document.

There is another dimension to this.  Let's say we want to remodel a space in a building. The first thing the architect will want to see is the existing floor plan and the other professionals will also want to see other aspects of the space.  Normally at this point the building owner will ask everyone involved in the planning to come over for a  walk-through the building.   That requires a lot of valuable time from very busy people.  With the virtual model that we can generate, most of this won't be necessary any more. The owner can ask a relatively junior person equipped with the backpack to walk the building to collect the data.  Then the resulting model can be put online and everyone can virtually walk through it.  For bidding on a big renovation project, a relatively junior operator does the walk through and puts it on the cloud and the construction companies who are interested in the project can virtually walk through it online for bidding purposes.

For a new building,  during the actual construction you can conduct a walk through every day and update visually what's going on.  Then if there are lawsuits in the future, you will have a permanent record of who did what and when.   You can show the legal folks where a component was supposed to go and where it was actually installed. It is also useful for monitoring progress, so you know who did what when to determine who gets paid and how much.   If there is a lawsuit and insurance companies get involved, you can step through the record to see who did what.  Then when you want to sell the finished building, the real estate folks can do a virtual walk through of the building with prospective customers.  Property managers and facility managers can use the virtual model for maintaining the building. For example, we talked about the energy performance audit.  Facility managers can use it to identify lights that need to be replaced with more efficient bulbs. Then they can put a virtual tag on the lights for the maintenance folks.  With a virtual model like this, owners, contractors, designers, and facility managers can all collaborate throughout the lifecycle of the building.

There seems to be a lot of areas where the technology you have developed could be used ?

There are quite a number of other application areas for the technology we have identified including historical building and monument recording, scan-to-BIM,  variance detection and virtual inspection, real estate valuations, building plans for first responders, change management and dispute resolution, and detecting thermal anomalies including water leaks and insulation deficiencies to name just a few.

GeoBuiz 2016: Geospatial and consumer devices, reality modeling, big data and analytics

Reality modeling

In the future, the deliverables at the end of a construction project will not only include the completed building itself, but also a digital "reality model" of the structure.  The digital model will include digital as-builts comprised of georeferenced point clouds captured by laser-scanning or meshes generated from images captured with a digital camera on a smartphone.  That was the prediction of Bhupinder Singh of Bentley at the GeoBuiz 2016 Summit in Bethesda.  This is a vision that is shared by other professionals in the construction industry including Ron Singh, Chief Surveyor at the Oregon Department of Transportation, who sees it as dramatically changing the construction process from surveying through design and construction to operations and maintenance.

Consumer devices and professionals

Consumer mobile devices will increasingly be the platform for many professional activities which currently require specialized devices and a PC computing platform.  Bryn Fosburgh of Trimble pointed out that while currently the location accuracy of GPS equipped smartphones is 10 meters, centimeter precision is surely coming because all that is required is a relatively simple antenna redesign.  Bryn sees a continuum in these devices between consumer and professional applications.   But while consumers tend to blindly accept whatever is available on their smartphones, professionals using these devices will have to be aware of the limits of the hardware, what it can be used for and what not.  For example, while smartphones in the future will be capable of centimeter accuracy in ideal locations, the accuracy will be less where there is a lot of scatter in the RF signals from the GPS satellites.  To ensure reliable results a professional surveyor relies on multiple sources including GPS linked to an earth station and traditional total station measurements.  Increasingly we will require other professionals to be involved.  Bryn reported a rising demand for photogrammetrists, as a result of the growing volume and availability of high resolution imagery which needs professional to help interpret.

Platform as a service

For virtually all of the world's big IT companies including IBM, Oracle, SAP, and Microsoft and the world's big content companies such as Digital Globe, the cloud is central to their vision of the future.  Together with others in the IT industry Xavier Lopez of Oracle sees the future in platform as a services (PaaS), but there are challenges.  A major challenge is integrating data in different databases including traditional SQL databases from Oracle and Microsoft and nonSQL databases such as Hadoop.  Another challenge is semantics.  Frequently different disciplines use different terms for the same thing.  I have blogged about the CB-NL project in the Netherlands aimed at solving the semantics problem in the construction industry.  Another big issue that Xavier identified is data fusion without co-location - how do you enable efficient processing of a logically federated database when the data resides in different locations ?

Monetizing the cloud

But perhaps the biggest challenge is finding an appropriate business model.  It is not completely transparent how the cloud is going to be successfully monetized, though the trend is toward some form of subscription or pay-per-use.   An interesting model of the successful application of the cloud in the construction industry is Textura, just acquired by Oracle.  Textura’s cloud serves as a place in which all of the participants in a project from the funding, through design, to construction can come together and access the shared assets necessary to complete the project.  Textura’s network includes about 85,000 general and subcontractors who using Textura to collaborate on construction projects.  Textura handles on average about 6,000 projects each month. The company’s shared database, which hosts project budgets, contractors, invoices, payments, change orders and compliance documents, resides in the cloud and is accessible to contract administrators, contractors, developers, subcontractors, consultants, title companies, and banks.

Big data and analytics

One of big drivers for the cloud is enabling the extraction of information from the huge volumes of data (most of which includes location) which are now available.  I have blogged about the nearly 100 petabytes of earth observation imagery available from Digital Globe.  According to Dr Richard W. Spinrad, Chief Scientist at the National Oceanic and Atmospheric Administration (NOAA) , NOAA is expecting to have 160 petabytes of archived geospatial data by 2020. 

As another example, Jeff Jonas of IBM described a publicly available database which includes 600 billion transactions per day collected by telephone companies.  This data, which includes location, can be mined (after being anonymized) for very interesting and commercially valuable information.  Jeff Jonas described some examples - using it to analyze where people spend most of their time (their "pattern of life"), how many people go into a store, and how long they remain in a store. 

Dipanshu Sharma, Founder & CEO, xAd illustrated live some of the things his company, which is in the location-based marketing business, does with this type of data.  xAd can customize ads to target people who are inside a store ("There's a great deal on aisle 5"), just outside one ("Come on in, we have great deals on umbrellas") or close to a competitor's store ("Checkout XXXX, our prices are lower than YYYY") - the words are mine.

In all of these example, geospatial data and technology are key to the applications.  Jeff Jonas foresees that the apps of the future on your smartphone will require geolocation and will be so compelling that just about everyone will enable GPS location on their smartphones.