LDAC2020 - 8th Linked Data in Architecture and Construction Workshop (17-19 June 2020)

The LDAC workshop series provides a focused overview on technical and applied research on the usage of semantic web, linked data and web of data technologies for architecture and construction (design, engineering, construction, operation, etc.). The workshop aims at gathering researchers, industry stakeholders, and standardization bodies of the broader Linked Building Data (LBD) community. The aim of the workshop is to present current developments, coordinate efforts, gather stakeholders, and elaborate use cases.


KEYNOTES

Three keynotes took place, on the afternoon of the 17th, and the morning of the 18th and 19th of June, with the following presentations:

    Wednesday 17 June (16:00 - 17:00)

  • The Socio-technical Phenomena of Data Integration and Knowledge Graphs
  • Juan Sequeda
  • Abstract: Data Integration has been an active area of computer science research for over two decades. A modern manifestation is as Knowledge Graphs which integrates not just data but also knowledge at scale. Tasks such as schema and ontology matching, data virtualization, etc., are fundamental in the data integration process. Research focus has been on studying this phenomena from a technical point of view (algorithms and systems) with the ultimate goal of automating the task of integrating data. In the process of applying scientific results to real world enterprise data integration scenarios to design and build Knowledge Graphs from enterprise databases, we have experienced numerous obstacles. In this talk, I will share insights about these obstacles. I will argue that we need to think outside of a technical box and further study the phenomena of data integration with a human-centric lens: from a socio-technical point of view.

    • Thursday 18 June (09:30 - 10:30)

  • The Smart Appliances REference Ontology (SAREF), its development and its applications
  • Laura Daniele
  • Abstract: In this talk I will take you into a journey that started in 2014, when the European Commission launched the first initiative to build a common ontology in close collaboration with the smart appliances industry, which resulted into the creation of the Smart Appliances REFerence ontology (SAREF). Six years later, SAREF is a series of technical specifications published by the European Telecommunication Standardization Institute (ETSI), consisting of a modular framework that comprises a generic core ontology for IoT and 10 domain-specific extensions, including SAREF for Energy, Buildings and Cities, which are of particular interest for the LDAC community. The SAREF framework is maintained and evolved by experts from several European organizations that successfully collaborate with each other and can count on the continuous support of ETSI and the European Commission. One of the latest supported initiatives is the development of an open portal for the SAREF community and industry stakeholders, so that they can contribute directly to the SAREF evolution. On the practical side, the recently started H2020 Interconnect Large Scale Pilot uses SAREF and its extensions as basis to enable interoperable solutions connecting smart homes, buildings and grids in various pilots located in seven different countries in Europe. During the talk I will share with you the lessons learned during this journey and the challenges ahead, addressing questions and curiosities, like what makes SAREF a successful story, how to keep an ontology relevant to the industry and its community of users, how to consistently maintain and evolve extensions in various domains, but also more specific topics for the LDAC community, such as how to model buildings and different domains related to buildings, how to link SAREF to other domains, and what are the challenges when concretely using the ontology to develop large scale real applications, especially when going across-domain like the Interconnect project aims to do when combining the different domains of smart homes, buildings and grids.

    • Friday 19 June (09:30 - 10:30)

  • History of the Semantic Web, and some words about the future of AI
  • Ali Intizar
  • Abstract: Due to the rapid advancements in the sensor technologies and IoT, we are witnessing a rapid growth in the use of sensors and relevant IoT applications. A very large number of sensors and IoT devices are in place in our surroundings which keep sensing dynamic contextual information. A true potential of the wide-spread of IoT devices can only be realized by designing and deploying a large number of smart IoT applications which can provide insights on the data collected from IoT devices and support decision making by converting raw sensor data into actionable knowledge. However, the process of getting value from sensor data streams and converting these raw sensor values into actionable knowledge requires extensive efforts from IoT application developers and domain experts. In this talk/tutorial, we will discuss various approaches for designing intelligent IoT applications and building real-time data analytics pipeline. We will present a common framework to design and build data analytics based IoT applications. We will discuss different challenges at each of the data processing layer of IoT data analytics pipeline and how a combination of modern technologies such as data analytics (machine/deep learning algorithms), semantic Web, Linked Data and AI can help address these challenges. A few example use-cases from smart cities, Industry 4.0 and construction domains with emphasis on digital twins will be presented in detail to demonstrate the utilisation of real-time IoT data analytics pipeline for building intelligent IoT applications.

PLENARY SESSIONS

These sessions, which take place after the introduction in the morning of the 17th and keynote on the 18th of June (wednesday and thursday), focus on current research and developments on linked building data, with the following presentations:

Wednesday 17 June (10:00 - 12:00)

  • From obXML to the OP Ontology: Developing a Semantic Model for Occupancy Profile
  • Serge Chavez-Feria, Giorgos Giannakis, Raúl García-Castro, and María Poveda-Villalón
  • Abstract: Building energy performance simulation is vital for predicting energy performance under specific conditions. Even though there exist several calculation methodologies that can provide precise simulation approximations, the accuracy of these tools is affected by the level of detail of its input data, being occupant behavior the main source of uncertainty in building energy performance results. Hence, initiatives such as the obXML schema have been developed recently to get a deeper understanding and to generate a more elaborated description of these data. Besides, the use of semantic web technologies can open opportunities to link behavioral patterns with other building information and related domains allowing interoperability and heterogeneous data integration. This paper presents the Occupancy Profile ontology, a semantic model based on obXML that allows the representation of occupant behaviors and actions inside building spaces. This paper also provides a catalogue of modelling decisions taken during the development and an example.
  • Linked Data for Smart Homes: Comparing RDF and Labeled Property Graphs
  • Alex J. A. Donkers, Dujuan Yang, and Nico Baken
  • Abstract: The need to integrate siloed data in the built environment led to a gaining interest in semantic web technologies in the Architecture, Engineering and Construction (AEC) sector. Especially for smart home developments, the integration of information about the building, users and Internet of Things (IoT) devices could be valuable. The Resource Description Framework (RDF) is the standard model for the semantic web, however, labeled property graphs (LPG) also proved to be effective in linking data. This research used the Open Smart Home Dataset and a dataset representing a kitchen to compare the two graph models both qualitatively and quantitatively. The comparison shows that native labeled property graphs are less complex and outperform the atomic RDF in complex graph traversals. However, RDF shows qualitative advantages for multi-domain and multi-stakeholder environments, such as the use of ontologies and HTTP URIs, making it a more stable interoperability format.
  • Towards defining Data Usage Restrictions in the Built Environment
  • Gonzalo Gil and Iker Esnaola-Gonzalez
  • Abstract: The building sector consumes about 40% of global energy, which is largely caused by buildings' operations. Research showed that providing users with detailed consumption and appliance-usage data may engage them in energy efficiency activities. To do so, data analytic services must be deployed by third parties. These services require a huge amount of data from the users in order to offer high quality services. Nevertheless, users are usually reluctant to share their data especially if its private. This is the case of those measurements performed inside the buildings, as they may be an indicative of the occupants' behaviour. In these scenarios, data sovereignty plays a key role because it provides to data owners a way to define data usage requirements and verify their compliance in a trustworthy way. In consequence, the owners' predisposition to exchange their data is expected to increase. This article takes a step forward in the development of a trustworthy ecosystem for the built environment where data sovereignty is provided. Specifically, an approach is proposed to formally specify data usage requirements for the built environment based on ontologies. Furthermore, this approach has been implemented in a real world home use case.
  • Design and integration of the project-specific ontology for data analytics support
  • Miloš Šipetić, Reinhard Jentsch, Judit Aizpuru, and Jan Kurzidim
  • Abstract: Data analytics projects often have access to little or no metadata associated with data collected and intended for processing, data analysis or machine learning. Even when metadata is available, it is usually poorly described, structured and linked to the corresponding data. In H2020 project InBetween, metadata description problem is approached by designing the project ontology on top of established ontologies BOT, SAREF and SAREF4BLDG. An approach for structuring and integrating all the collected data about occupants, their dwellings, appliances, sensors and locations into the project knowledge base is described. Building monitoring data, weather observation and weather forecast streams stored in SQL-based and time-series databases are semantically linked to the information about their respective source sensors stored in the knowledge base. Two data analysis use cases are discussed, along with improvements and benefits enabled by using the linked data approach. Discussed use cases are virtual occupancy sensor and electricity load profiling service. Linked data approach and use of common and standardized dictionaries help designing data analysis workflows that are easier to test, reuse and reason about.

    • Wednesday 17 June (14:00 - 16:00)

  • A GIS-based Ontology for Representing the Surrounding Environment of Buildings to Support Building Renovation
  • Maryam Daneshfar, Timo Hartmann, and Jochen Rabe
  • Abstract: This research focuses on developing an ontology for representing knowledge about the surrounding environment of a building in an urban context, considering the geospatial objects and processes such as built environment, vegetation, population and so on. The ontology can be useful to create a knowledge management system for different experts involved in the process of the building renovation, to extend the information and stretch the domain from the individual building to the environment. Knowledge about what entities and attributes to select is captured based on literature and investigating the pilot demonstration sites. Such an ontology can help to structure the surrounding data to support processes in different stages of the renovation. The final goal is to support planners in decision making process namely in site planning and pre-data collection phase, energy modeling, comfort analysis and so on to control cost and quality. Moreover, it can be valuable in further studies of integrating data of various sources for construction purposes.
  • Integration of BIM-related bridge information in an ontological knowledgebase
  • Al-Hakam Hamdan and Raimar J. Scherer
  • Abstract: Currently, utilizing digital representations of bridge constructions is still limited to geometry-based models with none or only little semantic data. Consequently, assessing these models requires the interpretation of external sources, e.g. relational databases or reports. Despite new approaches in the Building Information Modeling (BIM) domain such as the IFC-Bridge extension for the Industry Foundation Classes (IFC) try to provide models where geometric and semantic data are combined, a great proportion of information from current and future domains that are relevant for the bridge industry are not covered. For this reason, an extensible web ontology inspired from the Building Topology Ontology (BOT) for buildings has been developed, which functions as a core ontology for bridge representations and therefore covers all necessary general information used in this domain. The developed core ontology is interlinked with multiple domain specific bridge extensions in a bridge ontology framework that is applied on a test scenario. In this paper, the components of the bridge ontology framework are explained as well as the application on the test case. In addition, ontology alignments for BOT and ifcOWL are proposed as well as shapes for ontology validation. Furthermore, the functionality of a developed software prototype is described that generates the bridge ontology from a given IFC model.
  • Validation of IfcOWL datasets using SHACL
  • Sander Stolk and Kris McGlinn
  • Abstract: Standardisation is an important part of ensuring data interoperability. Industry Foundation Classes (IFC) is the current leading standard for BIM in the Architecture Engineering and Construction (AEC) industry and ifcOWL is a Resource Description Framework (RDF) representation of IFC, which enables the interlinking of IFC models with other building and building related data that are also represented using RDF, such as devices, sensor data, geolocation, etc. IFC has a complex schema, designed to support parametric modelling in AEC and adherence to this schema is required to support importing IFC models into popular CAD tools such as Autodesk and ArchiCAD. Therefore, for those wishing to create ifcOWL models which can then be imported into these tools, a process of validation of the output must be done. In this paper, we present a method for validating ifcOWL models using SHACL which can be reused by anyone generating ifcOWL models and which returns a report highlighting any issues identified. The method is tested to validate the outputs of a conversion of geospatial data into IFC using a declarative mapping approach called R2RML.
  • ifcOWL-DfMA a new ontology for the offsite construction domain
  • Edlira Vakaj, Franco Cheung, Abdel-Rahman Tawil, Panagiotis Patlakas, Kudirat Alyania
  • Abstract: Architecture, Engineering and Construction (AEC) is a fragmented industry dealing with heterogeneous data formats coming from different domains. Building Information Modelling (BIM) is one of the most important efforts to manage information collaboratively within the AEC industry. The Industry Foundation Classes (IFC) can be used as a data format to achieve data exchange between diverse software applications in a BIM process. However, EXPRESS, the native IFC modelling language, is not a logic-based language and lacks formal semantics; consequently, intelligent tools cannot engage with this information directly. The advantage of using Semantic Web Technologies to overcome these challenges has been recognised by the AEC community and the ifcOWL ontology, which transforms the IFC schema to a Web Ontology Language (OWL) representation, is now a de facto standard. Even though the ifcOWL ontology is very extensive, there is a lack of detailed knowledge representation in terms of process and sub-processes explaining Design for Manufacturing and Assembly (DfMA) for offsite construction, and also a lack of knowledge on how life cycle cost and carbon emissions are incurred, which is essential for evaluation of alternative DfMA design options. In this article we present a new ontology named ifcOWL-DfMA as a new domain specific module for ifcOWL with the aim of representing offsite construction domain terminology and relationships in a machine-interpretable format. This ontology will play the role of a core vocabulary for the DfMA design management and can be used in many scenarios such as life cycle cost estimation. To demonstrate the usage of ifcOWL-DfMA ontology a production line of wall panels is presented. We evaluate our approach by querying the wall panel production model about information on activity sequence, cost estimation per activity and also the direct material cost. This ultimately enable us to evaluate the overall product from the system.

Thursday 18 June (11:00 - 12:30)

  • Pattern-based access control in a decentralised collaboration environment
  • Jeroen Werbrouck, Ruben Taelman, Ruben Verborgh, Pieter Pauwels, Jakob Beetz, and Erik Mannens
  • Abstract: As the building industry is rapidly catching up with digital advancements, and Web technologies grow in both maturity and security, a data- and Web-based construction practice comes within reach. In such an environment, private project information and open online data can be combined to allow cross-domain interoperability at data level, using Semantic Web technologies. As construction projects often feature complex and temporary networks of stakeholder firms and their employees, a property-based access control mechanism is necessary to enable a flexible and automated management of distributed building projects. In this article, we propose a method to facilitate such mechanism using existing Web technologies: RDF, SHACL, WebIDs, nanopublications and the Linked Data Platform. The proposed method will be illustrated with an extension of a custom nodeJS Solid server. The potential of the Solid ecosystem has been put forward earlier as a basis for a Linked Data-based Common Data Environment: its decentralised setup, connection of both RDF and non-RDF resources and fine-grained access control mechanisms are considered an apt foundation to manage distributed building data.
  • Common Data Environments for the Information Container for linked Document Delivery
  • Madhumitha Senthilvel, Jyrki Oraskari, and Jakob Beetz
  • Abstract: Unstructured and poorly managed information is a major cause of time delays in construction projects. Availability of relevant information at the required time has a considerable impact on decision making in the project's lifecycle. Multi-models containers, and a more recent approach, Information container for linked document delivery (ICDD), aim to facilitate construction data management and sharing information. Containers help in structuring and linking of heterogeneous data. Such container models are relevant in and align with Common Data Environments (CDE) which facilitate a centralised environment for managing both information and services. Presently, three approaches fit in the vision of CDE: the DIN SPEC 91391-2, OpenCDE-API which is loosely based on DIN SPEC 91391-2, and the W3C Linked Data Platform (LDP), a generic data container-based approach to managing linked data graphs online. In this paper, we investigate how ICDD, an approach to link information, can be represented using the above three frameworks. A comparison is developed between the three approaches based on a sample use-case. The required conversion steps are analysed, and the limitations of the mapping are evaluated.
  • Linking BIM and GIS Standard Ontologies with Linked Data
  • Elio Hbeich and Ana Roxin
  • Abstract: Following the analysis of existing BIM and GIS standards, formats, differences in the interpretations of the underlying concepts have been identified. Still, in each of the two considered domains several ontologies have been defined for these terms without seeking an alignment among their definitions. With this scope in mind, this article presents several mappings expressed by means of explicit semantic links between GIS concepts (as present in the related ontologies for the ISO 191XX standard family) and BIM concepts (as represented in the IFC standard ISO 16739:2018). Such semantic mappings are defined in order to ensure a knowledge continuum between both domains, thus enabling seamless reasoning in application contexts spanning over them e.g. urban contexts.

INDUSTRY TRACK

On the afternoon of June 18 (thursday) and morning of June 19 (friday) we will focus on recent developments and proposals from the industry, with the following presentations:

Session 1 (14:00 - 17:30)

  • General, using linked data solutions to integrate the construction industry
  • John Egan (BIMLauncher)
  • Abstract: An independent report suggests Cost of double data entry incurs 73% productivity loss, 70% slowed info workflows, 62% frequency of data entry errors (Source: connectedaec.com) when exchanging information between common data environments. Shaped by requirements set out in international information management standards, CDEs are faced with both technical and governance based challenges to connect data sources. Linked data can be a means to create an environment for multiple project management systems to lift the barriers and allow seamless flow of information that can fundamentally change the way we work together on projects as an industry. With the growth and continuous disintegration of platforms and Applications I will discuss what it will take for linked data solutions to “integrate” with the construction industry.
  • General, data communication between disparate design applications using Neo4J and GraphQL
  • Will Reynolds (Hoare Lea)
  • Abstract: A big inefficiency in building design lies in communicating data between disparate design applications. This talk will explore how a graph database (specifically Neo4j), and a query language (GraphQL), can provide a common source of truth from which these applications can easily push and pull the data they need. To further present the viability of the concept we’ll consider how a cloud hosted API server (GRANDstack) can provide a single, scalable, endpoint for building data throughout its lifecycle. We’ll explore how graph data adds huge value by capturing important relationships between these components which are a key step towards achieving building data utopia.
  • Norms/standards, BIM standards ISO 23386 and ISO 23387
  • Espen Schulze (Cobuilder)
  • Abstract: These standards are needed to build trust in the content of information exchanges used in construction platforms and software to connect towards Manufacturers and to take a step towards Circular Economy, Green deal whilst ensuring implementations of regulatory frameworks such as Construction Products Regulation and Low Volt Directive. This methodology will provide core business semantics that Linked data can use to provide an eruption of our industry.
  • On integration between manufacturer data and building models using linked data in compliance with ISO 23386
  • Nicolas Bus (CSTB)
  • Abstract: Manufacturers usually produce data on products using corporates data structures and vocabularies. How have them sharing a common frame to help customer comparing products ? how to integrate manufacturer data with building models ? The ISO 23386 standard defines common process and meta-data to implement and manage a properties thesaurus called data dictionary. In this presentation we describe an approach to implement ISO 23386 with benefits from Linked Data. Do you we need new vocabularies ? or can we just align ISO 23386 ontology on existing Link Data vocabularies ? In first application we show how building experts can use a Linked Data oriented Data Dictionnaries compliant with ISO23386 at design stage. How can they deal with the various international classifications ? We show this application through the business case of a French company delivering certificates on building. The second application concern building owners using products databases for retrofitting. How to link products with properties thesaurus? How to integrates data from Building elements and manufacturer products ? This use case came from the BIM4REN European project.
  • On the FireGraph knowledge graph that utilizes LBD ontologies
  • Bart van Leeuwen (Netage B.V.)
  • Abstract: Netage B.V. has created FireGraph a knowledge graph for the fire service where we have used the current version of BOT to link into building information from various sources to improve the information position of fire services. Based on open standards FireGraph provides fire services with a information architecture on top of which they can build their information products either for operations or for preparation. Due to the scattered nature of data available to the fire service integrating these sources is in general cumbersome, with the knowledge graph approach we have enabled fire services to get a complete overview of the available information regardless of what the datasource is. To integrate the various sources of building information we have utilized the Ontologies which are under construction in the Linked Building Data working group.
  • Smart buildings, Google's own real estate, brick, haystack
  • Charbel Kaed (Google)
  • Abstract: Google manages its own real estate by collecting data from different buildings to better manage them and optimize their operations. In this work, we present our internal platform which collects data from BMS systems and our Digital Buildings ontology, based on HayStack and Brick in order to semantically represent 130 buildings from Google’s buildings in the California Bay Area region. The ontology proposed in this work is designed with our subject matter expert in HVAC mechanical systems.

Session 2 (10:30 - 12:30)

  • Data integration, FM, practical use case from schiphol airport, asset data management
  • Jan Voskuil (Taxonic)
  • Abstract: Schiphol (the international airport) has outsourced to 6 Main Contractors (MCs) the management of its 70,000 assets, ranging from doors and elevators to landing guidance systems and Airplane Docking Stations. Each MC is the single source of truth regarding asset information. With each update, the MC sends a message containing the full description of the pertinent asset to SNBV’s CDV, the “central data hub”, a high-performance native RDF graph database with application layer. The data are published in near-real time to the Asset Viewer. Data stewards use the CDV for data governance and quality monitoring, based on business and validation rules in SHACL. In the future, apps will be created that use the auto-generated GraphQL API. We at Taxonic are now in the process of productizing the solution. We have partnered up with Macomi, specialized in state-of-the-art asset analytics, for predictive maintenance, line-of-sight monitoring, long-term scenario evaluation and the like. Our vision is that asset data must be FAIR. In this talk, I will talk about this vision, on how our first implementation at Schiphol worked out, and on what we learned from it. And what we are planning. With an eye on challenges related to OTLs and ontologies.
  • Infrastructure, roads, Bjørnafjorden Open Live Centre (BOLC) and the V440 ontology
  • Lars Wikström (Triona AB)
  • Abstract: The presentation will give an overview of the state in a large infrastructure project in Norway, E39 Coastal Highway, where a concept called Bjørnafjorden Open Live Centre (BOLC), an interactive web-portal for live project information, is introduced. This portal shall manage data from many different sources and data formats. To manage all these data types in a structured way, a well-suited classification system is key. The Norwegian Road Authority has a classification system defined in their handbook for bridge registration; V440, used for planning, design, construction, maintenance and operations. The project develops an ontology representing V440 based on LD/SW-technologies. The development is led by the Technical Room of BuildingSMART Norway. The resulting ontology, including support for linking to other relevant ontologies, will be available for all stakeholders. Currently, the project is in a phase where PoC:s (Proof-of-concept) are being developed together with software companies. When the BOLC web-portal is realized, then this will be a full scaled pilot project for how to apply the LD/SW techniques for classification of data from a variety of domains, sources and formats. This will be something new and unique in this scale and really show how open standards can support advanced open BIM projects.
  • Smart buildings, integration between a maintenance management system and a SCADA
  • Eduardo Gilabert (Tekniker)
  • Abstract: BIM (Building Information Modelling) is a collaborative working methodology for the creation and management of constructive projects. Its objective is the centralization of all the project information, generated by all the agents, into a digital representation model. However, as it happens more often than not, the overuse of buzzwords or trendy terms diffuse true sense and BIM is no exception. The BIM shouldn’t be confused with IFC files, which are mainly restricted to exchange geometrical building data. Instead, a BIM must be understood as a central model which includes, links and integrates information that covers the whole building life cycle. Certainly, the Semantic Technologies can contribute to the integration of these heterogeneous data sources, and allow easy and intuitive ways to rapidly browse, query and use this information. In the context of the NIERBIMO project, the integration between a CMMS (Computerized Maintenance Management System) and a SCADA containing building data is aimed via the use of BIM, with the main target of making a more efficient and sustainable use of energy, operation and maintenance in buildings starting from the planning phase.
  • On the need for "more lenses to view the same data", connecting structured and unstructured data and making it available using structures defined in ISO 81346
  • Philipp Dohmen (Amberg Group)
  • Abstract: Data-driven analytics is at the core of global businesses today and the construction industry of tomorrow. But any good data analytics strategy requires a capacity to quickly obtain valuable insights from large amounts of data coming from diverse data sources. For real benefits it is necessary to connect structured and unstructured Data to gain better insights for planning, construction and operations. Todays problem is the essential need of complete different data structures per stage. The datastructure designed in ifc is for planning but doesn’t solve the needs of a construction site or operations.