Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 ifcOWL-DfMA a new ontology for the offsite construc- tion domain Edlira Vakaj Kalemi1, Franco Cheung1, Abdel-Rahman Tawil1, Panagiotis Patlakas1 Kudirat Alyania2 1 Birmingham City University, United Kingdom 2 University of East London, United Kingdom edlira.vakaj@bcu.ac.uk Abstract. Architecture, Engineering and Construction (AEC) is a fragmented in- dustry 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 Foun- dation Classes (IFC) can be used as a data format to achieve data exchange be- tween diverse software applications in a BIM process. 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 explain- ing Design for Manufacturing and Assembly (DfMA) for offsite construction, and also a lack of knowledge on how product and productivity measurement such as production costs and durations 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 vocab- ulary 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 ontol- ogy a production line of wall panels is presented. We evaluate our approach by querying the wall panel production model about information such as activity se- quence, cost estimation per activity and also the direct material cost. This ulti- mately enable users to evaluate the overall product from the system. Keywords: Offsite Construction, IFC, Ontologies, Linked Open Data Copyright © LDAC2020 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). 105 Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 1 Introduction The Architecture, Engineering and Construction (AEC) sector has been criticised as low in productivity as compared with that of other sector’s, e.g. the productivity of manufacturing, automotive, and aerospace sectors. One holistic approach to improve productivity is the application of Design for Manufacture and Assembly (DfMA). DfMA, first developed for product design, aims to improve production so that products produced are consumed by the manufacturing process as quickly as possible with the least amount of waste and redundant works. In practice, it involves a continuous eval- uation of the manufacture and assembly processes by designers. It is now widely ac- cepted within the AEC that one major direction to improve productivity is to move the production activities offsite [1]. The application of DfMA thus enable designers to con- sider alternative offsite production approaches with automation in mind. The use of Building Information Modelling (BIM) in building projects offers opportu- nities to extract properties and data of a building easily but there is generally a lack of attention to data collected during the construction process. Typically, process-related data in BIM are only used for scheduling purposes. The actual use of process data for informing manufacturing or off-site decision-making is limited. This paper proposes a semantic approach for linking process data with life cycle costs and carbon emissions to give an accurate production costs and carbon footprint. The estimations from the semantic knowledge-based system will give an objective measure to inform designers in evaluating DfMA options. Semantic Technologies and Linked Open Data have been broadly used in the domains of AEC. The usage of these technologies is driven by the need to operate with hetero- geneous data formats from different sources and domains, support data interoperability, flexible data exchange and distributed data management. An extensive literature review conducted by Pauwels et al. [5] has emphasised the crucial role semantic technologies and logic-based applications play in systems that require the integration of information from multiple application areas. The standard schema for the exchange of BIM data is IFC [11]. It has a strong focus on 3D geometry [11] and is modelled using EXPRESS [10]. Semantic Technologies where applied to implement a direct mapping of IFC EXPRESS schema to ifcOWL ontology [6]. IFC schema and ifcOWL ontology con- cepts of design differs from those used for DfMA as the latter is production led, focus- ing on the manufacture and assembly process. One example would be product classifi- cation in DfMA design in which products come with details of sub-assemblies, gener- ally are under-represented in the IFC schema. The proposed ifcOWL-DfMA ontology aims to provide the AEC community with a vocabulary of commonly understood concepts and relationships to represent the domain of offsite construction, as well as a means to publish linked open DfMA data. This is achieved by contributing to the development of domain knowledge that handles inter- disciplinary information exchange among different participants during the life-cycle of 106 Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 design for manufacturing. In addition it provides a basis for future development of smart tools that will be able to provide answers for practical scenarios. 2 Domain Knowledge and State of the art 2.1 Design for Manufacturing and Assembly Traditionally, design and construction are separated with the relevant responsibilities assigned to different parties. The role of a contractor is mainly an integrator that focuses on the delivery of buildings with little attention to the potential benefits of factory pro- duction. The call for improvement of the AEC sector in terms of productivity and prod- uct performance has led to a change in some market segments of the sector to consider alternative approaches to design and construct buildings. DfMA is a design approach that is composed of two parts: design for manufacture (DfM) and design for assembly (DfA). Through engineering a building design – often, a standardised design, the goal of DfMA is to minimise waste and redundant operations. Examples of specific targets for DfM are the selection of materials that minimise wastage and handling, optimising processes and sub-processes, optimising parts and systems fulfill tolerance require- ments, and those for DfA are minimising number of modules for assembly and opti- mizing assembly. In practice, it is a continuous task of reviewing, evaluating, rational- ising, standardising and optimising the functionality, producibility, handling and fixing of design. The task is very knowledge intense and complex, and requires input from experts of various disciplines - some of them such as production engineers are not traditionally a part of the building design team. As the knowledge is not readily accessible, there is a need to systemise the knowledge to enable the evaluation of building design by indi- vidual discipline owners. The current approach for evaluation relies heavily on either heuristic (“rule of thumb”) or high-level estimations with little effort spent on under- standing how processes and sub-processes are related and interacted. For instance, an estimated cost – as a measure for rationalizing or optimizing - is calculated based largely on historical high level per unit cost without taking into account on how cost actually incurred. This is problematic as the economy of off-site manufacturing, a core element of DfMA, is process-driven and can only be evaluated properly if the estimate reflects the cost implications of processes. For instance, the cost for a static production process, i.e. the use of mainly labour for production would be different from an auto- mated production, i.e. the use of mainly machine or robot for production. The knowledge however is not typically kept in the system of the current status quo. The argument that construction processes and sub-processes are premature to consider in the design stage in traditional approach does not apply if DfMA is to be adopted as building design is based on standardised design. Standardised design makes product, production and assembly data to be kept and reuse in a more efficient manner. System- ising knowledge of product, production and assembly for DfMA through creating an 107 Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 accurate representation of the relationships of the process and sub-processes can auto- matically generate estimates of productivity and performance matrics such as produc- tion cost, life cycle cost, duration or CO2 emissions. 2.2 Building Information Modeling Building Information Modeling is a digital process for the representation and pro- cessing of all information relevant to the Building Life Cycle (BLC). Typically, the foundation of a BIM process will be a three-dimensional (3D) model of the architectural design, detailing the positioning and dimensions of a buildings components (walls, win- dows, doors etc.) and facilitating the inclusion of non-physical building features such as building cost, accessibility, safety, security and sustainability [3]. In a BIM model not only the geometric features are included but also the semantic attributes are in- cluded and the associated properties [5]. BIM is an intelligent model-based process that connects AEC professionals so they can design, build and operate buildings more effi- ciently. BIM is also used for creating data for infrastructure associated with physical and functional characteristics. BIM projects are implemented from the start as either closed or open models. The latter uses the Industry Foundation Classes (IFC) [11], a standardised platform-neutral schema, for data exchange. An IFC data model in prac- tice focuses on building geometry representation in the design stage. In the work pre- sented here, the Open BIM approach is the assumed adoption. In the UK, BIM implementation is defined according to different levels of maturity starting from BIM Level 0 to Level 3 - a cloud-based implementation where data from different domains can be integrated seamlessly without any data loss [2]. The Semantic Technologies and Linked Data principles proposed can also play a very important role in achieving Level 3 BIM. 2.3 IFC and ifcOWL Ontology IFC files represent BIM components using the EXPRESS modelling language [10]. Using IFC data and instance serialization formats, BIM data can be exchanged between heterogeneous software applications. A basic overview of IFC hierarchy is given in Figure 1. However, IFC is not a web-compliant one, therefore there is a requirement to use semantic standards and technologies [11] like the Web Ontology Language (OWL) for which the Linked Data standards was proposed. Initially OWL was integrated with IFC [11] to produce ifcOWL ontology. Later, a direct mapping of EXPRESS schema to OWL [4] was introduced and implemented in the current version of ifcOWL ontol- ogy. The ifcOWL is now under buildingSMART [8] International, where it eventually became a part of the ISO 16739 standard [7]. 108 Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 Figure 1Part of the hierarchy of classes in ifcOWL ontology. The ifcOWL ontology is an extensive ontology. In the latest version, i.e. IFC4, con- sists of 1293 classes and 1572 object properties. This makes reasoning and management very hard and inefficient, and inevitability, increases the need to develop separate mod- ules based on the core IFC modules. The proposal to implement a modular ifcOWL ontology was proposed by [15] and has started to be adopted by different authors. Even more recently the need for modularity and extensibility was explicitly from the authors a [14] when they introduced the BOT - Building Topology Ontology. 3 ifcOWL-DfMA Ontology Development The aim of the ifcOWL-DfMA ontology is to present an ontology that defines the key terms and relationships present in the DfMA approach to building design, while simul- taneously acting as an extension of the ifcOWL ontology, in order to maintain compli- ance with core IFC concepts. 3.1 Ontology Development Methodology As illustrated in Figure 2, the first step taken to design ifcOWL-DfMA ontology was conducting a literature review in terms of: i) existing ontologies designed of IFC where ifcOWL was identified and analyzed; ii) existing ontologies for offsite construction, DfMA and related domain; iii) general DfMA and related domains literature review in order to extract the main concepts and relation of the domain. The literature review confirmed that there is no existing ontology that represents offsite construction and life cycle assessment with the DfMA approach. As a second step, a set of competency questions was drafted based of the guide for developing an ontology from Stanford University [16].The competency questions have 109 Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 guided the discussion with the stakeholders and experts involved including architects, production engineers, structural engineers, steel supplier, client and cost consultant on one to one interviews and group discussions. An iterative approach was adopted to the ontology design process to reflect the feedback from the experts and improve the on- tology. Competency Validation Questions Literature Expert Interview Review Figure 2 Methodology used to develop ifcOWL-DfMA ontology 3.2 Overview of ifcOWL-DfMA Ontology ifcOWL-DfMA define a model of categories within the offsite manufacturing Universe of Discourse (UoD), plus sufficient knowledge about those categories to allow for them to be reasoned upon and classified automatically. Our aim is to use ifcOWL-DfMA as a COmmon REference, or CORE model for offsite manufacturing. The proposed on- tological model is language independent, using the broader term ‘terminology’ for a semantic model linked to the offsite manufacturing domain. A high level schema (upper ifcOWL-DfMA ontology) is a prerequisite for categori- sation and integration, as illustrated in Figure 3: • Fits closely with building standards especially in applications for design and manufacturing assembly or in the retrieval and classification of ifcOWL-DfMA concepts. • Sufficiently general to be used in different applications for decision support and interoperability. • Formally defined in OWL Description Logic (DL) and can be considered a gen- eral-purpose modelling language for offsite manufacturing. • Supports OWL-DL reasoners to allow for core ifcOWL-DfM concepts to be combined to create new descriptions of classes and instances constructed accord- ing to constraints implemented within the ontology. • Support intuitive and practical collaboration between different groups, being eas- ily understood and application independent. 110 Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 OWL Thing Activity_Cost_Drivers Cost Activity_Cate- DfMA_Produc- gory tion_Process Platform Produc- Activities Resources tion_Cost Time Location Production_Ac- Labour Direct_Cost Indirect_Cost tivities Supporting_Activ- Material Transport ities rdfs:subClassOf Over- head Plant Figure 3 A high level schema of ifcOWL-DfMA ontology In ifcOWL-DfMA, the primary breakdown is into: • DfMA_Production_Process, defines both production and supporting activi- ties • Resources, defines labour, material overhead and plant • Activities, defines production activities (e.g., gladding assembly line auto- mation, frame assembly line) and resources (e.g., labour, material and com- ponent, overhead etc.) • Modality, defines platform, time, location and transport, representing a heter- ogeneous grouping for usage in associations with production processes and activities. A secondary structure is superimposed over the primary, aiming to capture DfMA production process, activities and resources. Table 1 shows ifcOWL-DfMA taxonomy of major elementary categories associated with the production process. The category labelled DfMA_Production_Process represents the disjunction of two main categories, activities and resources which can be observed. As an ontology for offsite design and manufacturing, ifcOWL-DfMA further divides production processes into production activities e.g., Cladding_Asssembly_Line, Frame_Assembly_Line and supporting activities such as loading, packaging and trans- porting. For example, a cladding assembly line is an automated activity that is defined as a production activity which begins only after frame assembly line is completed and consumes some labour. Cladding_Assembly_Line_Automated and beginsAfter only Frame_Assembly_Line Cladding_Assembly_Line_Automated and consumeLabour some Labour Loading isSubClassOf Suppoting_Activity and consumeLabour some Labour 111 Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 Resources are further divided into MaterialandComponents (e.g., Direct_Matrial and Packaging_Material), Overhead (e.g., Cleaning, Security) and plant (e.g., Mova- ble_Tools, Static_Tools). MaterialandComponents is used to group Direct_Material to- gether including external wall cladding, internal wall cladding, wall finishing, wall fix- ing and wall framing. Table 1. ifcOWL-DfMA taxonomy of major elementary categories Entity Example DfMA_Production_Process Activity Production_Actvity Cladding_Assembly_Line Adhesive_Station T31_Feed_Adhesive, T32_Dispense_Adhesive Briquette_Apply_Sta- T35_Place_Briquettes tion T34_Feeding_Briquettes Briquette_Load_Station … Frame_Assembly_Line T13_Return_Conveyor Conveying_Station T5_Rivet_Joints, T6_Move_Frame_to_Lift, Frame_Riveting_Station T7_Lift_Frame Frame_Transfer_Station T9_Transfer_Frame FrameBeam_Load_Sta- T1_Deliver_Pallets, T2_Se- tion lect_and_Load_Beam, T3_Clamp_Beam … Supporting_Activity … Resource Labour Direct_Labour Casual_Labour hasLabourHrRate "9.0"^^xsd:double Semi-Skilled_Operative hasLabourHrRate "13.0"^^xsd:double workingOnActivity T1_Deliver_Pallets workingOnActivity T2_Select_and_Load_Beam MaterialandComponent Direct_Material EXT_Wall_Cladding INT_Wall_Cladding Wall_Finishing Overhead Plant Subcategories share common characteristic from which a single constraint may be inherited, but are otherwise disjoint and heterogeneous. In addition, ifcOWL-DfMA recognises, UnitsOfMeasures such as miles, kilometres, metre, kilogram, minute, cur- rency etc. which are used as part of quantities. 112 Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 ifcOWL-DfMA Attribute Hierarchy The taxonomy of ‘attributes’ (or ‘semantic link types’) is influenced by and supports the outlined category taxonomy. The primary distinction here is between object and data properties. While data properties (e.g., hasUnitRate, hasLabourHrRate, hasCount) describe what kind of values a triple with the property should have by relating individ- uals to literal values (e.g., strings, numbers, datetimes, etc.), object properties (e.g., be- ginsBefore, beginsAfter, isComponentPartOf) relates concepts together to define rela- tionships across concepts. Rules in the form of the Semantic Web Rules Language (SWRL) are used to provide more powerful deductive reasoning capabilities than OWL alone. For example, the rule below determined the cost of labour for an activity by multiplying the processing time with the labour hourly rate. Activities(?a), Labour(?s), hasLabourHrRate(?s, ?r), hasProcessTime(?a, ?p), workingOnActivity(?s, ?a), multiply(?result, ?p, ?r) -> hasActivityCost(?a, ?result) 3.3 Alignments with ifcOWL ontology As mentioned in previous sections ifcOWL-DfMA ontology is developed inde- pendently from the ifcOWL ontology but is aligned with it such that every dfma:Build- ing is an IfcBuilding and every dfma:Product is an IfcProduct. ifc: Building ifc: Element rdfs:subClassOf dfma:Product rdfs:subClassOf dfma:SubAssembly dfma:Building dfma:OffsiteSystem dfma:Components dfma:Assembly Figure 4 Alignment between ifcOWL-DfMA ontology and ifcOWL ontology. 113 Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 Developing ifcOWL-DfMA as a separate domain of the existing ifcOWL ontology was a conscious choice. Naturally, many aspects of a completed DfMA project, such as the building geometry or the material properties, fit ifcOWL concepts and can be repre- sented accordingly. However, as a process with roots in industrial engineering, DfMA engages more with procedural and optimisation aspects, and introduces concepts, such as “assembly” or “sub-assembly”, with different semantics from current BIM and con- struction technology practice. As such, a separate ontological domain was considered necessary in order to avoid semantic and ontological conflicts, as well as to implement DfMA concepts appropriately. Ideally, ifcOWL-DfMA will be able to facilitate a two- way conversation: enable AEC practitioners to apply DfMA design concepts in a BIM workflow, while simultaneously acting as an introduction to the DfMA concept to BIM- literate AEC practitioners. At the same time, the need for a separate domain suggests that there are some limita- tions to the current ifcOWL ontology. Attempting to capture all possible aspects of a building in a single hierarchical ontology, mapped to a super-schema, has innate limi- tations and lacks the flexibility to accommodate different design concepts. DfMA is a characteristic case study on that: future innovative philosophies and practices are likely to face similar challenges in BIM implementation. 4 Using ifcOWL-DfMA ontology in practice The ifcOWL-DfMA ontology is applied on the production of a wall panel system for a house design using DfMA. The application captures its production process and adopts a manufacturing costing approach, namely Activity-Based Costing (ABC) to classify cost data. The process-based costing method measures the activity costs of cost objects (i.e. various cost centres for wall panels) attempting to give accurate and traceable cost information. Decision makers are thus presented with more in-depth information that encourages corrective actions. For instance, it allows users to identify cost drivers of an off-site wall panel production such as factory rent and production volume. A separate process mapping exercise for DfMA production is carried out and a process map for proposed off-site production line for DfMA house wall panels has been produced. The wall panels are modeled by describing their attributes such as the components that com- pose a wall panel but also the production line detailed in terms of activities carried to produce a wall panel as illustrated in Figure 5. All activities are connected with each other by keeping track of which activity should perform first(hasStartingActivity) and which activity takes place next(hasNextActivity) or in parallel. Further on, the knowledge represented in the ontology is used to estimate cost (hasDirectCost, hasMaterialCost, hasActivityCost etc.) per each activity and over- all cost of producing one product in this case a wall panel. By estimating the cost per each activity the designer can get insights in which activity are occurring overhead costs and optimise their design if possible. 114 Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 Figure 5 Modeling wall panel instantiation Apart from costing, Table 2 gives some example queries that a designer might possibly ask regarding the DfMA house composed of 32 wall panels to the instantiated ontology, which are the estimates of potential productivity and performance matrices. The queries are expressed in SWRL or SQWRL and the reasoning is made by Pellet reasoner. Table 2. Example queries in SWRL/SQWRL and respected results. Question Query(SWRL/SQWRL) Result in Protégé Comment Q1. What is labor Semi-Skilled_Operative(?s) ^ Activities(?a) "1.17"^:SSO_1 The result displayed cost for each ^ workingOnActivity(?s, ?a) ^ hasProcess- "2.6" ^:SSO_4 shows the labour cost of semi-skilled oper- Time(?a, ?p) ^ hasLabourHrRate(?s, ?r) ^ "5.85"^:SSO_6 building a single wall ative working on swrlb:multiply(?result, ?p, ?r) -> "1.4689"^:SSO_3 panel automatically on each activity of sqwrl:sum(?result) ^ sqwrl:select(?s) "3.237"^:SSO_5 the production line with the wall panel five operatives support- production? ing different activities done by the robots to form the panel. Q2. What is total Prod- "3075.81918" This result aggre- direct material uct(LSF_3BED_01_LHS)^hasDirectMateri gates the cost of each cost for producing alCost(LSF_3BED_01_LHS, ?m) -> components/materials panel? sqwrl:select(LSF_3BED_01_LHS) ^ used in building up the sqwrl:sum(?m) panel of a semi-detached house (with identity 01) 115 Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 Q3. What are the Product(?p) ^ hasComponentPart(?p, ?Com- mmS mmSS This results show the components of ponent) -> sqwrl:select(?Component) mmStud mmHT components used in WallPanel01? mmHT mmBT building the wall panel mmCS mmStud 01 which includes mmBT mmBT Studs, Head Track, Base mmBT mmFS Track, Cripple studs of mmStud different sizes. Q4. What is the Product(LSF_3BED_01_LHS) ^hasStart- LSF_3BED_01_LH This result displays starting and up- ingActivity(LSF_3BED_01_LHS, ?Star- S, T1_Deliver_Pal- the sequence of activi- coming activity tActivity) ^ hasNextActivity(?StartActivity, lets, T2_Selectand- ties carried out by oper- for producing ?NextActivity)->sqwrl:se- LoadBeam atives in building the LSF_3BED_01_L lect(LSF_3BED_01_LHS, ?StartActivity, wall panel with identity HS wall panel? ?NextActivity,) 01 5 Conclusion and Discussions This article proposes a new domain specific ontology ifcOWL-DfMA ontology which expands ifcOWL ontology as a separate module deriving from core element of IFC. The ifcOWL-DfMA ontology is however on the early versions of development and further improvements can be done. In order to ensure interoperability this ontology is rooted in the de-facto standard ontology for IFC (ifcOWL) and follows the Linked Data principles. To address the complexities that ifcOWL has, the World Wide Web Consortium (W3C) Linked Building Data Community Group is standardizing the Building Topology Ontology (BOT) [14] that will interlink different domain specific ontologies more efficiently and when this is needed. BOT uses Linked Data approach to describe the buildings by only using the fundamental properties and if more detailing are required the linking with other relevant ontologies is enabled. This is the direction that ifcOWL-DfMA is planning to take after wider evaluation with the community of interest. Acknowledgments This research was supported by an Innovate UK funded project “Collaborative Knowledge- Based DfMA approach to build cost efficient, low impact and high perfor- mance houses”. The authors would like to thank the participating industrial partners of the research project, particularly Walsall Housing Group and Northmill Associates for their contribution in this study. 6 References 1. Barbosa, F., Woetzel, J., Mischke, J., Ribeirinho, M.J., Sridhar, M., Parsons, M., Bertram, N. and Brown, S., 2017. Reinventing construction through a productivity revolution. McKinsey Global Institute,(https://goo. gl/1Nqqf8)(Feb. 06, 2017). 2. British Standard Institute (2013) PAS 1192-2: Specification for Information Management for the Capital/Delivery Phase of Construction Projects Using Building Information Model- ling, British Standard Institute. 116 Proceedings of the 8th Linked Data in Architecture and Construction Workshop - LDAC2020 3. Taylor, J., Bernstein, P.: Paradigm Trajectories of Building Information Modeling Practice in Project Networks. Journal of Management in Engineering - J MANAGEENG 25(2009). https://doi.org/10.1061/(ASCE)0742-597X(2009)25:2(69) 4. Li, N., Li, Q., Liu, Y.-S., Lu, W., & Wang, W. (2020). BIMSeek++: Retrieving BIM com- ponents using similarity measurement of attributes. Computers in Industry, 116, 103186. https://doi.org/10.1016/J.COMPIND.2020.103186 5. Pauwels, P., Zhang, S., & Lee, Y. C. (2017). Semantic web technologies in AEC industry: A literature overview. Automation in Construction, 73, 145–165. 6. Pieter Pauwels, Walter Terkaj, EXPRESS to OWL for construction industry: Towards a recommendable and usable ifcOWL ontology, Automation in Construction, Volume 63, 2016, Pages 100-133, ISSN 0926-5805, https://doi.org/10.1016/j.autcon.2015.12.003. 7. Pauwels, P., Krijnen, T., Terkaj, W., & Beetz, J. (2017). Enhancing the ifcOWL ontology with an alternative representation for geometric data. Automation in Construction, 80, 77– 94. https://doi.org/10.1016/j.autcon.2017.03.001 8. BuildingSMART, https://www.buildingsmart.org/, Accessed: February 2020 9. ISO 16739, Industry Foundation Classes (IFC) for data sharing in the construction and fa- cility management industries, 2013, International Standardisation Organisation, Geneva, Switzerland. 10. ISO 10303-11, Industrial automation systems and integration – Product data representation and exchange – Part 11: Description methods: The EXPRESS language reference manual, 2004, International Standardisation Organisation, Geneva, Switzerland. 11. J. Domingue, D. Fensel and J.A. Hendler, Handbook of semantic web technologies, Springer, 2011. 12. J. Beetz, J. van Leeuwen and B. de Vries, IfcOWL: A case of transforming EXPRESS sche- mas into ontologies, Artificial Intelligence for Engineering Design, Analysis and Manufac- turing 23(01) (2008), 89. 13. P. Hitzler, M. Krötzsch, B. Parsia, P.F. Patel-Schneider, S. Rudolph, OWL 2 web ontology language primer (second edition) — W3C recommendation 11 December 2012, W3C rec- ommendation, available online: http://www.w3.org/TR/2012/REC-owl2-primer-20121211/ (Last accessed on February 2020) 14. M. H. Rasmussen, M. Lefrancois, G. F. Schneider, P. Pauwels, BOT: the Building Topology Ontology of the W3C Linked Building Data Group, Semantic Web (2020) 1–21Accepted, Last accessed: 27 March 2020. URL http://www.semantic-web-journal.net/sys- tem/files/swj2279.pdf 15. Terkaj, W., & Pauwels, P. (2017). A method to generate a modular ifcOWL ontology. CEUR Workshop Proceedings. 16. Krötzsch, M., Simanˇ, F., & Horrocks, I. (2013). A Description Logic Primer ∗ arXiv : 1201 . 4089v3 [ cs . AI ] 3 Jun 2013. 17. N. F. Noy, D. L. McGuinness, Ontology Development 101: A Guide to Creating Your First Ontology, Tech. Rep. SMI-2001-0880, Stanford University, Stanford, USA, last accessed: 27 March 2020 (2001). URL https://protege.stanford.edu/publications/ontology_develop- ment/ontology101.pdf 117