Entrant details
Role or Job Title on the Project
Project Manager
Employer
Trimble Solutions Sandvika
Sandvika
Norway
Employer Role
Technology or Software Development Company
Are you or your employer a member of buildingSMART?
Yes - Chapter Member
Submission details
Submitting Party Company Name
Norwegian Public Roads Administration
Submitting Party Company Location
Oslo, Norway
Submitting Party Role on Project
Project sponsor
Submitting Party Company Website
Full Project Name
Machine Readable Norwegian Classification Manual for Bridge Registrations - V440
Project Location (Country)
Norway
Project Objectives
Today the Norwegian bridge registration manual, V440, is issued by the Norwegian Public Roads Administration as a PDF-document. The project objective was to make the classification structure described in V440 available on an open machine readable format to allow for software vendors to integrate this classification structure into their software. Bridge designers using the software can then produce IFC-files containing the V440 classification system.
The V440 classification ontology is developed using Linked Data and Semantic Web technology.
Prototyping the integration of the ontology in software solutions was an important part of the project. Four software vendors participated in this prototyping.
openBIM Achievements
The V440-ontology is produced using open Linked Data and Semantic Web (LD/SW) technology. LD/SW is based on technology standardised by ISO and World Wide Web Consortium. The formats used in LD/SW are fully open and should be considered as an integrated part of openBIM.
During the prototyping phase, the V440 ontology was linked to the ifcOWL ontology for IFC 4.3, demonstrating how easy it is with LD/SW technology to link national classification systems with IFC Bridge definitions and make this available in software solutions.
The resulting classified bridge models were transported to other software solutions for registration using the IFC-format.
openBIM used
IFC 2x3, IFC4, ifcOWL
openBIM or open standards used other than those listed above
Ontologies defined by Linked Data /Semantic Web (LD/SW) technology are based on a series of open standards defined and maintenanced by ISO and World Wide Web Consortium (W3C). Some of the most important of these open standards, in addition to the "basic www-standards", are:
- RDF: Resource Description Framework
- RDFS: Resource Description Framework Schema
- SPARQL: RDF Query Language
- OWL: Web Ontology Language
Software used
In this project we used to different groups of software:
- Software for defining and publishing the ontology
- Software for prototyping and registration of the produced IFC-files.
In addition GitHub is used as the project collaboration platform for information, Q&As and for exchanging ontologies and linking rule sets.
Software used for defining and publishing the ontology:
- MS Excel
- Protégé
- TopBraid Composer
- GraphDB
Software used for prototyping and registration:
- Areo
- Quadri
- Tekla Structures
- StreamBIM
- Solibri
Strategic Alignment
LD/SW technology has been an integrated part of the technology platform within buildingSMART for many years, leading to the development of ifcOWL e.g.. But LD/SW has not drawn much attention outside the group of people already interested. The focus has been file-transfer.
With the modernisation initiative of IFC, the focus has shifted to support more web-service based architecture. Now the LD/SW technology is more in focus.
This project is an excellent showcase to demonstrate how easy it is for software vendors to integrate LD/SW ontologies into their software, link the ontology with ifcOWL and produce valid IFC-files.
Highlights
Important highlights we have achieved in this project are:
- Showed the benefit of moving «digital paper» based (PDF-format) classification systems to an open machine readable format.
- Showed how easy it is for software vendors to integrate well documented ontologies into their software to produce valid IFC-models.
- Showed the benefit of using the LD/SW technology to link different ontologies together with linking rule sets.
- Developed a lean and simple ontology modelling process based on Excel spread sheet for domain experts and Top Braid Composer for the LD/SW experts.
- Made very good demo showcases for the use of LD/SW technology in four different software solutions.
- Teached interested softwre vendors the basic concepts of LD/SW and how to implement this technology in existing software packages.
- The links to the V440 ontologies have now been published at the official Norwegian Public Roads Administration (NPRA) site for manual V440. NPRA now expect the industry to use these ontologies in their daily work. https://www.vegvesen.no/fag/teknologi/bruer/bruhandboker/v440-maskinlesbar-bim
Project Website
Project Address
This project is managed by the Norwegian chapter of buildingSMART International, Technical Room.
The Technical Room can be reach on this e-mail address:
teknisk@buildingsmart.no
Project Type
Civil
Size of Project
- The project started in August 2019 and ends in August 2020.
- The total budget for the project is 920 000,- NOK
- Four software vendors participated actively in the project by prototyping. They included the developed ontology into their software
- Another 8** software companies participated in workshops and demos
Detailed description of the project
Project initiative
The initiative to launch this project came from one of the larger infrastructure projects owned by the Norwegian public Roads Administration (NPRA); E39 - Bjørnafjorden fjord crossing. Bjørnafjorden project aims to build a long floating bridge with an elevated section, crossing the wide fjord, Bjørnafjorden, south of Bergen on the western coast of Norway, https://www.vegvesen.no/en/roads/Roads+and+bridges/Road+projects/e39coastalhighwayroute/news/a-floating-bridge-is-the-chosen-concept-for-crossing-bjornafjorden
Included in this project is an open live centre to survey all the ongoing activities an sensor data relevant for the project. This solution is called Bjørnafjorden Open Live Centre (BOLC).
BOLC needs a common classification system to align the incoming data-streams. NPRA decided to apply the bridge classification defined in the bridge registration manual called V440.
Due to the high probability of revisions of the classification system during the timespan of the project, NPRA understood that they needed a “loose coupling” of the classification system from the applied software packages to allow for a frequent revision of the classification structure independent of update frequency of software packages.
NPRA decided to test if the Linked Data /Semantic Web (LD/SW) technology was a good solution for handling the classification structure. NPRA contacted buildingSMART Norway, Technical Room for assistance to run an innovative technology project engaging as many relevant software vendors as possible.
The work performed in the project
The project was organized in three phases:
Phase one
As the first step in phase one, the domain experts in NPRA moved the information from the pdf-formatted V440 manual to a spread sheet in a structure that easily could be imported into TopBraid Composer to define the ontology. Some scripting (SPIN) was applied to the ontology in the TopBraid Composer to make the ontology follow the best practises for ontology modelling as defined by the ongoing work in CEN/TC442 Semantic Modelling and Linking Standard e.g..
The V440 ontology was divided into two ontologies, one representing the classification structure itself and the other representing the related properties.
Phase two
In phase two, NPRA and the project team did quality assurance revisions of the ontologies and then published them at a server site provided by NPRA. In addition a SPARQL endpoint was published, allowing the software vendors to use the querying language SPARQL to access the classification structure. GraphDB was used as the triplestore to provide the SPARQL endpoint.
A GitHub site was established, containing all the relevant information needed for the participants in phase three of the project, including ”getting started” material and the possibility to ask questions. The wiki part of this site is very well developed.
Here are the relevant web-addresses:
http://rdf.vegdata.no/V440/v440-owl (bridge classification system described in V440)
http://rdf.vegdata.no/V440/v440-brudata-owl (related properties described in V440)
http://rdfspatial.vegdata.no:7200/repositories/V440 (SPARQL endpoint)
https://github.com/buildingsmart-norway/maskinlesbar_v440 (GitHub site)
Phase three
In phase three we invited software vendors to implement the use of the published ontologies in their software.
We followed a prototyping process called [openLAB : Interoperate] adapted by buildingSMART Norway. This is a "workshop" that last for several months where the participating software vendors can develop their solutions and exchange information and experiences.
We set up this [openLAB : Interoperate] process to last for about three months from May until June 2020. We were able to get 10 companies to sign up for [openLAB : Interoperate]. Out of these 10 companies, 4 companies (BuildingPoint Scandinavia, Rendra, Trimble solutions Sandvika and Areo) ended up with implementing the use of the V440 ontologies in their software solutions.
We started the [openLAB : Interoperate] with a kick-off meeting, introducing the participants to the classification implementation challenges and we arranged a small crash course in LD/SW technology.
During the period of three months, we organized two workshops where the participants could exchange information and demo their solutions. We also introduced some new challenges in these workshops; linking the V400 ontologies to ifcOWL and exporting the resulting BIM-models as valid classified IFC-files e.g.
At the end of the [openLAB : Interoperate] period we organized a closing seminar, inviting all relevant organisations and companies in Norway. The participants demoed their developed solutions and report their experiences. We arranged a competition among the software vendors in the project as well. Areo was the winner.
About 30 companies and organisations attended the closing seminar. This is the streaming from the seminar: https://www.youtube.com/watch?v=Ds3CzR7X4-c&t=5640s
Project results
The result of phase one and two of this project is documented on the GitHub site and with the links to the ontologies given above. The production process proved to be very efficient. The domain specific classification considerations were done by domain experts in their familiar Excel tool. The ontology modelling was then done by LD/SW experts in TopBraid Composer.
There were four software vendors implementing the V440-classification. They all found the provided documentation on GitHub easy to follow. The implementation was fast. About one day from the introduction until the first classification was shown in their software.
The vendors tested both the use of the SPARQL endpoint for dynamic reading of the classification and they tested to download the turtle files from web-links to use these files in a more integrated way in the solutions.
The vendors exported the resulting classified BIM-models as IFC-files and showed them in Solibri.
Some of the vendors experimented with making their own linking rule sets to link together ifcOWL and V441 (Norwegian bridge damage classification manual) with V440.
The prototype results are documented both in the OpenBIM Supporting Evidence attachement and in the general attachments.
Dissemination
An important part of the project was to engage software vendors in the construction industry to start implementing LD/SW technology, so we carried out several activities to promote the use of LD/SW.
We arranged a breakfast meeting after phase two of the project where we invited external speakers in addition to presenting our own project and the explaining the benefits of using LD/SW. More than 50 different companies and organizations participating. https://www.youtube.com/watch?v=vSesWj667Ck&feature=youtu.be
We made a video for promoting the concept of LD/SW: https://www.youtube.com/watch?v=Ms8frmt2jD4
We presented the project at several conferences such as infra BIM Open in Tampere and LDAC (Linked Data in Architecture and Construction) e.g..
Detailed description of openBIM on the project
Here openBIM is defined both as the IFC-formats listed in the award application-schema and the different open formats and standards supporting Linked Data and Semantic Web.
The main “technical element” in this project is the V440-ontology itself. This ontology is based on the open formats RDF (Resource Description Framework), RDFS (Resource Description Framework Schema) and OWL (Web Ontology Language) with their underlying open standards and formats.
We made two V440-onlotologies and made them public available here:
http://rdf.vegdata.no/V440/v440-owl
http://rdf.vegdata.no/V440/v440-brudata-owl
The first ontology contains the bridge classification system described in V440
The latter ontology contains all the connected properties described in V440
These ontologies are modelled according to the Semantic Modelling and Linking Standard, currently under development by CEN/TC442. The structure of the ontologies follows ISO 12006-2 as much as possible, given the defined classification structure in the V440 manual.
A Tripplestore based on GraphDB is established here: http://rdfspatial.vegdata.no:7200/, including a SPARQL endpoint: http://rdfspatial.vegdata.no:7200/repositories/V440.
A GitHub site has been established for the project: https://github.com/buildingsmart-norway/maskinlesbar_v440
On this site we have included some linking rule sets showing how the developed V440 ontologies can be linked to the ifcOWL ontology for IFC 4.3: https://github.com/buildingsmart-norway/maskinlesbar_v440/tree/master/ontologi/Linking%20Rule%20Sets/LRS_v440-IFC
On the GitHub site we also included some IFC-files containing bridges. These files could be a starting point for the software vendors that did not have any bridge models available for prototyping: https://github.com/buildingsmart-norway/maskinlesbar_v440/tree/master/modellfiler
Some of these files already included some classification to demonstrate how the resulting exported IFC-files from the prototype implementations could be.
Benefits from using openBIM
To use of openBIM and other open standards is the very essence of this project. The project used only open formats provided by ISO and W3C when defining the ontologies and linking rule sets. In addition, the formatting of the contents at the published endpoints used open formats/syntax like Turtle and SPARQL e.g.. The resulting classified BIM-models were exported from the prototype editions of the software to the validation software as IFC 2x3 and/or IFC4 files.
This project clearly showed how fast software vendors can implement the use of Linked Data and Semantic Web (LD/SW) ontologies in their software when the ontologies are published in an open, standardized and well documented format.
The LD/SW technology allows for the use of external sources for model- and/or classification definitions separated from the software and the BIM-models themselves. These model definitions and/or classification ontologies can then be maintained by the organisations responsible for the classification and/or model definition (typically national bodies responsible for a certain built environment domain e.g.).
The updating frequency and versioning of these ontologies will be independent of the updating frequency of the applied software. An updated/revised classification structure e.g., can easily be applied on existing BIM-models with the use of supporting software without the need of any software updates.
By applying the IFC-format for the resulting classified BIM-models when exporting them from the prototyping software, it was super-fast to check if the classification of the bridge models were correct by using a model checker like Solibri e.g..
"We were able to innovate using openBIM."
We have defined this project as an innovation project. Linked Data and Sematic Web (LD/SW) technology is not used much in the construction and building industry today, even though the technology has been in use for many years in other industries.
A main aspect of this project was to teach interested software vendors how to apply the LD/SW-technology in their software solutions so that external sources of model information, like classification e.g., easily can be applied to the BIM-models developed by the software. For these software vendors this project was a real eye opener and an innovative new way of handling external requirements to the BIM-models.
Several larger consultancy companies and national bodies responsible for different domains of the built environment in Norway, also participated as stake holders in the project. They joined in at workshops and demos. For these organisations, the result of this project really was considered as an innovative new way of organizing the national and/or project specific standardization work for BIM-models.
"We were able to identify where we need openBIM to develop further."
This project was set up to test and prototype the use of Linked Data and Semantic Web (LD/SW) in existing software. By opening up for LD/SW as the technology platform for openBIM, there are almost no limits for the possibilities you can achieve by linking and combining different ontologies.
We made some simple linking rule sets ontologies for combining the V440 ontology with the ifcOWL ontology for IFC 4.3. This approach should be investigated further to link the national V440 bridge classification system with the international IFC Bridge object type definitions.
There is a manual issued by the Norwegian Public Roads Administration for classification of bridge damages, called V441. An obvious next step would be to make an ontology of this manual as well, following the development process used in this project. Then a linking rule set could be developed to link these two ontologies (v440 and V441) together.
A nice feature to adapt when using LD/SW, is to link illustrations to the different elements in the ontology. These illustrations that can then be applied by the software systems as help information for the users. The Norwegian manuals have a rich set of illustrations that could have been made available as linked illustrations into the ontologies.
The ifcOWL for IFC 4.3 is more or less automatically created based on the IFC 4.3 schema. This leads to an ontology that is not defined in an optimal way and is not following all the best practices for ontology definitions. The ifcOWL for IFC 4.3 could be revised to be more in accordance with best practises.
BIM Uses were defined on the project
I agree to be contacted about the project BIM uses outside of this awards program.
Stakeholders
Norwegian Public Roads Administration, Norway,
https://www.vegvesen.no, Sponsor, Elin Dalen-Rasmussen
Triona AB, Borlänge, Sverige,
http://triona.se, Modelling the ontology, Lars Wikström
MOK-SEE AS, Oslo, Norway,
https://www.linkedin.com/company/mok-see/, Technical expert, Sigve Martin Pettersen
buildingSMART, Norway, Technical Room, Oslo, Norway,
https://buildingsmart.no/, Project management, Jan Erik Hoel
Areo, Oslo, Norway,
http://areo.io/, Prototyping, Hans Kristian Grani
Trimble Solutions Sandvika, Sandvika, Norway,
https://www.novapoint.com, Prototyping, Reidun Hansen
BuildingPoint Scandinavia, Oslo, Norway,
https://buildingpoint-scandinavia.com, Prototyping, Espen Skogsrud
Rendra, Oslo, Norway,
https://streambim.com, Prototyping, Ole Kristian Kvarsvik
