Interoperability methods for Infrastructure Design and Information

Transcription

1 Interoperability methods for Infrastructure Design and Information Systems Nazereh Nejatbakhsh Esfahani Technische Universität Dresden, Institute of Construction Informatics Abstract: There are many attempts to establish interoperability between Infrastructure Data Models and Geo-Information Systems (GIS). Organizations like buildingsmart and Open Geospatial Consortium (OGC) deal with the guidelines. LandXML is used for information exchange standardization in infrastructure planning in many countries. To facilitate exchange of data across multiple GIS systems OGC has promoted the CityGML standards. LandXML is suitable for exchange of surfaces, alignments and profiles and can pretty handle geometric / parametric road design. CityGML is strong in converging heterogenic descriptive models. Yet it doesn t cover the detailed functional aspects of transportation network. With the respect to the aims of the author, this paper explores the proper interoperability methods for the envisioned purposes of the work. To study these methods this paper provides complementary information regarding the requirements of the interoperability and elaborates the exploitation of each method for these requirements. Keywords: Interoperability, buildingsmart, OGC, LandXML, CityGML 1. Motivation The infrastructure design concerns strongly with the positioning of the physical elements of the design objects according to standards and constraints. The basic objectives in geometric design are to optimize efficiency and safety while minimizing cost and environmental damage. Geometric design also affects an emerging fifth objective called livability, including providing access to employment, schools, businesses and residences, accommodate a range of travel modes such as walking, bicycling, transit, and automobiles, and minimizing fuel use, emissions and environmental damage (Federal Highway Administration, 2012). These facts have called for applying Information Modeling (IM) in infrastructure which creates value in designing transport infrastructure, management and control of the entire project and increases profits and optimal results. 169

2 Historically infrastructure projects highly concern the location intelligence and hence geospatial data. In the early stages before the road is designed, GIS plays an important role in the selection of sites and integration of data throughout the process. Moreover, planning and implementation of infrastructure projects can take many years which depletes valuable resources and results in significant delays and costly deadlocks. This is especially true as communities struggle to engage all stakeholders in the development process due to the long and sometimes tedious approval process and the incompatibility of data and technology stacks used in various stages of the aforementioned project lifecycle. Here the project transitions back to GIS for the maintenance phase after IM is used for design and construction. Therefore both GIS and IM solutions support these projects on various scales and in all stages, including planning, building, management and maintenance (Abukhater, 2013). In reality these workflows are mostly not linked together well, which brings the development to a crawl. This delay can cost the community a significant amount of resources and lost opportunities. Main objective of the work is to investigate integration approaches between Infrastructure Design and GIS. This means to feed the right information for a more effective decision making to avoid duplications, parallel work or unnecessary recurring of planning. According to Esfahani et al. (2012) it is not the right approach to interoperate any kind of data domain in one comprehensive platform like IFC. The proper solution could be a system for management of the information and model logistics as well as the interdependencies among application models. Organizations such as buildingsmart and OGC are dealing with the guidelines. LandXML has been chosen for standardization of information exchange in road; railroad and water planning in most of the countries. Mentioning that a lot of current limitations as well as the data loss in export / import between applications are software related, De Nier R. (2013) describes pros and contras of the LandXML in a comparison to the CityGML as follows: Is well suitable for exchange of surfaces, alignments and profiles (+) Can pretty handle geometric / parametric road and railway design (+) Is not appropriate for exchange of the 3D models (-) Fails in dealing with semantic information (-) To facilitate the exchange of data across multiple GIS systems and indirectly across multiple views of the built environment the OGC has promoted the CityGML standards which in the case on infrastructure planning: 170

3 Is strong in converging heterogenic descriptive models (+) Uses generic modules for transportation networks (+) Does not cover the detailed functional aspects of transportation network models (-) Has a very abstract definition of the transportation network and does not contain explicit description of transportation objects (-) 2. State of Art There are many data formats in infrastructure domain, mainly due to the diversity of labor and the organization of the planning process as well as the construction industry, such as exclusive identity of the structures, project teams, construction, high proportion of the design in the overall services, no mass production, no prefabrication, high regulation, etc. Therefore unlike the wide application of business informatics in the automotive industry and the production chain, it is not possible, to apply the knowledge of economic computer sciences like SAP for the information management of the infrastructure domain. Existing proprietary design management and construction software will continue to produce and process the data in these formats. These applications have spread widely, are expensive and highly specialized, so that it isn t possible to simply introduce a new even multipurpose software. Besides these existing data formats are derived from different domains such as construction site, structure, engineering services, operations, etc. The constantly demand increase for cross-domain management of the project tasks, indicates the necessity of crossdomain data like geodata and infrastructure. So far there are several ways to integrate this data: 1. Schema integration Product / project data models integration in schema level is a very powerful and practical methodology to integrate different structures on spatial objects and representation in order to allow the reuse of existing data sources in the sense of schematic interoperability. To develop GIS in a conventional manner, first, the vocabulary of requirements analysis should be created to identify what data is needed through application requirement analysis. After that, the newly acquired data from various heterogeneous data sources can be integrated into a single, uniform data store. 171

4 Advantage: data all in one format, and possibly even in one database; eg IFC, CIS/2 Disadvantage: always new domains which are not included, eg geodata in IFC Although the data models are principally expandable but practically it does not work due to standardization, involved committees, discussions, etc. 2. Ontology Approach In the integrated systems the data from one format (eg LandXML ) get transformed through database or ontology into the second system (eg GIS). Guarino (1998) addresses that it would be convenient to agree on a single top level ontology rather than relying on agreements based on the intersection of different ontologies (Jang and Kim, 2007). Advantage: cross-domain links and a query language can be created Disadvantage: these data are no longer interchangeable Therefore the challenge is to make this data interoperable. 3. Methodology The OGC defines interoperability as the capability to communicate, execute programs, or transfer data among various functional units in a manner that requires the user to have little or no knowledge of the unique characteristics of those units (ISO ) Ontology Interoperability Methods 1. Metadata Interoperability Marine Metadata Interoperability describes on its website metadata interoperability as the ability of two or more information systems to exchange metadata with minimal loss of information. An important aspect is that interoperable metadata can be used by computer systems, in contrast to metadata that is designed to be read and understood by a person. Having interoperable metadata allows multiple systems to work with the same set of data and metadata. It ensures metadata records associated with one resource can be accessed, accurately interpreted and subsequently used by a system or integrated with metadata records associated with other resources. For example, interoperable metadata allows: tools, such as drawing systems, to easily and accurately import data. people to move geospatial datasets between various GIS systems. 172

5 a dataset to be searched and found through multiple catalogs without the provider having to implement multiple sets of metadata. Nowadays, most organizations in charge of cataloguing geographic metadata (in accordance with standards like CSDGM (Federal Geographic Data Committee (FGDC), 1998) or CEN/TC 287 prenv (European Committee for Standardization (CEN), 1998)) aim at migrating towards the international ISO standard. Apart from that, they are usually asked to provide a more generic description of their resources (Nogueras-Iso, 2004). Different institutions and projects will inevitably develop customized metadata templates and files, which must conform to essential standards, and include effective labeling, to ensure interoperability. This appeals in organizations, such as local governments, that have entities or departments that independently collect and manage spatial data (eg roads, pipes, surveys, land records, administrative boundaries). Simultaneously, many of the functions of a local government require these data sets to be integrated. A connecting technology coupled with an integrating machine such as GIS can efficiently support this need. As result various layers of information can be dynamically queried and integrated, while at the same time the custodians of the data can maintain this information in a distributed computing environment (Esri Inc. ). Once equipped with the standard, a technical tool can be implemented to translate between these customized metadata files (crosswalks). The existence of a vocabulary, such as a dictionary and / or thesauri can also aid in the development of interoperable metadata. 2. Semantic Interoperability Semantics concerns the study of meanings. Semantic interoperability refers to the ability of computer systems to transmit data with unambiguous, shared meaning. It is a requirement to enable machine computable logic, inferencing, knowledge discovery, and data federation between information systems (Mencarini and Melegari, 2011). NCOIC (2008) describes semantic interoperability to concern not just the data packaging (syntax), but the simultaneous transmission of the meaning with the data (semantics). This is accomplished by adding data about the data (metadata), linking each data element to a controlled, shared vocabulary. The meaning of the data is transmitted with the data itself, in one selfdescribing information package that is independent of any information system. It is this shared vocabulary, and its associated links to an ontology, which provides the foundation and capability of machine interpretation, inferencing, and logic. 173

6 Ontologies are considered to be an adequate methodology to support semantic interoperability in heterogeneous information systems including geospatial information systems since ontologies can provide a common basis for semantic mapping between information communities (Jang and Kim, 2007). Ontologies are explicit formal specifications of domains of discourse on the premise that ontologies facilitate knowledge sharing and reuse (Noy and Musen, 2003). Guarino (1998) elaborates further the notion of ontology as a logical theory accounting for the intended meaning of a formal vocabulary, i.e., its ontological commitment to a particular conceptualization of the world. He significantly contributed to the domain of ontology research by proposing the four levels of ontologies, which are: top-level ontology, domain ontology, task ontology, and application ontology. 3. Model Driven Interoperability Model Driven Interoperability (MDI) is a methodological framework, providing a conceptual and technical support to make interoperable enterprises using ontologies and semantic annotations, following Model Driven Development (MDD) principles (Humm et al., 2005; Mohagheghi et al., 2009). As shown in figure 1 the three main ideas of MDI approach are: Interoperability is achieved at Business, Knowledge, Application and Data levels. The main idea is to follow a Model Driven Engineering approach. Therefore, it is promoted a systematic use of models as primary engineering artifacts throughout the engineering life cycle combined with both domain specific modeling languages and transformation engines and generators. The use of ontologies and semantic annotations is needed in order to perform model transformation from enterprise level to code level. Figure 1: MDI among Enterprises at various data levels (Source: Wkinterop, 2012) In the field of Infrastructure, there are some attempts in semantic integration of model systems. OpenINFRA is an international initiative under the roof of BuildingSMART which aims 174

7 at improving interoperability for planning and realization of infrastructure facilities, such as roads, bridges and tunnels. By LandXML there are some attempts to preserve data integrity when importing and exporting data between two dissimilar systems. These are guidelines and not rules as to how to implement a system that supports LandXML data interoperability. Geographic Markup Language (GML) is an XML-based language developed by the OGC, an international organization founded in 1994 to promote interoperability of GIS Multi model for infrastructure domain Figure 2 describes that in a multi model (MM) the heterogeneous data models are bundled together keeping their original format into a container a so called link model where the elements of data model will be linked together. The interlinking of the elements to their metadata will remain untouched. The original data, their metadata and the link model will be saved in and exchanged through a container (eg. ZIP File). Figure 2: Multi Model Container (Source: Mefisto Projekt, Scherer et al.) Thus, interchangeable cross-domain information sharing environment results, which allows: Further use of existing software No need to create new data formats Obtain cross-domain information The core in a link model is how to interlink the data and how to analyze the linked data models to obtain cross-domain information. Scherer et al. have developed an MM (Scherer and Schapke, 2011) which is equipped with an interlinking engine and a query language (Fuchs, 2012) and a graphic which shows the interlinking, exchange, and cross domain filter. 175

8 The MM is extendable for other engineering domains like infrastructure. It inputs only certain data formats, and refers semantic metadata which are of exact understanding and usage for the information systems, eg railway, highway, tunnel, bridge, etc. This extension then provides a new set of filters which allows simpler queries for engineering level, eg radius of a curve, cut and fill, minimum and maximum length of the structures etc. where the alternatives come from infrastructure domain such as the LandXML and the spatial data from GIS domain such as CityGML. Therefore for the purposes of this work there are the following possibilities: 1. Use the existing generic multipurpose MM for infrastructure design and GIS domains 2. Specialize an MM for Infrastructure Domain 3. Embed one data model in the other one For infrastructure domain MM functions as to filter an element from infrastructure design model such as road by a defined criteria; then filter an element eg terrain from the GIS model by a specified criteria and at last to generate a link between the two filtered elements. The user must specify the criteria and filters. A customization on MM for infrastructure projects provides such specialized filter criteria and rules. Considering that: MM is a volatile object. It is created principally for special cross-model / cross-domain work task. The links could have totally different meanings in different tasks. Access to the databases is done perfectly MM is indeed not able to transform data from one format to another Existing data models will be re-used in their original software Existing as well as newly developed applications, which are intended to solve cross-domain problems, can use the MM engine MM is quite appropriate, if there are many different stakeholders, labors, types of tasks, questions and queries involved with the heterogeneous data. Talking about one specified task like evaluation or optimizing design using spatial data, it is more worthwhile to develop a task based algorithm which is much more specific. However in regard of data exchange one should rather combine the existing data formats than to introduce a new format with a huge effort. In the regard of data, metadata and semantic exchange both approaches follow the same concept but unlike ontology approach, MM is a cross domain approach which makes it applicable for most of the domains if not all. Both systems handle just pre specified geometries which results in very limited visualization environments. More importantly both approaches lack 176

9 in handle the real time changes on elements and dynamic data structure due to being based on predefined ready to use data models. MM takes the advantage of being cross domain and handling queries and filters and interrelates different data models together, yet is not developed to enable establishing spatial links, which is a bed rock for geodata and of course infrastructure design. Table 1 shows this comparison briefly: Table 1: Comparison between ontology and multi model based interoperability methods for Infrastructure and Spatial Information Data Models Criteria Ontology Multi Model Data Exchange, Semantic and Metadata + + Cross Domain - + Geometry - - Visualization - / Compatible with other BIMs / Extendable + + Schema Documentation + + Complexity - + Query and Filter - + Spatial Query and Spatial Filter - - Real Time / Dynamic (Analytical) Data Structure - - Yet, engineering design is not the only application field for GIS in the infrastructure domain. Smart technologies, smart heating systems, smart cars and smart highways (studio roosegaarde, 2011) with the goal to make roads more sustainable and interactive by using light, energy and road signs that automatically adapt to the traffic situation and weather condition, open a new world of application for GIS in infrastructure domain. Besides, there are many attempts since 2009 to use solar panels instead of asphalt. The solar panels would generate electricity, which would in turn be fed into the grid. This makes oil conserve twice: Electric cars could be charged with the energy produced by the panels, and the panels would replace the use of asphalt, the production of which requires petroleum. Moreover, Solar Roadways are heated and equipped with integrated LED screens, which act not only as street markings, but can also show warnings directly on the road (Breunig et al., 2012). According to the experiences, it is obvious that various algorithms are under development for different application fields such as spatial information, controlling, energy efficiency, etc. But it is pointless to develop many complex new formats and algorithms and which makes the interoperability and interchanging of the data much difficult. MM is here a great asset. 177

10 4. Conclusion Ontology based interoperability like semantic interoperability between infrastructure design and GIS, gives computer systems the possibility to transmit design and geodata with unambiguous, shared meaning. Semantic interoperability enables machine computable logic, inferencing, knowledge discovery, and data federation between application systems. MM is a general approach a link model which takes the heterogeneous data from their original format, interlinks them together and filters them in accordance to task based queries. It is quite appropriate, if there are many different stakeholders, labors, types of tasks, questions and queries involved with the heterogeneous data. The data is re-usable in its generic software. Talking about one specified task such as evaluating or optimizing infrastructure design using spatial data, it is more worthwhile to develop a task based algorithm, or an adjusted much more specific MM. These algorithms are supposed to handle spatial linking and geometry matching as basic interoperability criteria for these data models. Yet it should be taken into account that Infrastructure Domain is much wider than just infrastructure design and GIS, but these two are the two major schemas of the domain. Literature Abukhater A. (2013) GIS meets BIM: Delivering End-to-End Solutions for Government and Infrastructure. Pitney Bowes Publications. May 31, 2013 Breunig M. et al. (2012) Towards 3D Geoinformatics and Computational Civil Engineering Support for Cooperative Tracks Planning. Knowing to manage the territory, protect the environment, evaluate the heritage. FIG Working Week 2012, May 6-10, 2012, Rome, Italy Esfahani N.N, Scherer R.-J and Balder R. (2012) Traffic Infrastructure Design and Geo-Information Systems: a case of Interoperability. ework and ebisoness in Architecture, Engineering and Construction vol 9: Federal Highway Administration (2012) The Role of FHWA Programs in Livability: State of the Practice Summary. Fuchs S. (2012) Eine Abfragesprache für Multimodelle. Forum Bauinformatik 2012 Humm B., Schreier U. and Siedersleben J. (2005) Model-Driven Development Hot Spots in Business Information Systems. ECMDA-FA 2005, LNCS 3748: Guarino N. (1998) Formal Ontology and Information Systems. Proceedings of FOIS 98. June 3-15, 1998, Trento, Italy Jang S.-G. and Kim T. J. (2007) Semantically Interoperable Geospatial Information Processing: A Review and A Canonical Model Approach. 10th Association Geographic Information Laboratories Europe (AGILE). International Conference on Geographic Information Science, May 8-11, 2007, Aalborg, Denmark 178

11 Mencarini M. and Melegari A. (2011) Semantic Technology as a Key Enabler for Semantic Interoperability: How Semantic Technology Can Help Semantic Interoperability in Connecting the Dots. Semantic and Domain-based Interoperability. Information Systems Technology (IST) Panel Symposium. November 7 8, 2011, Oslo, Norway Mohagheghi P., Dehlen V. and Neple T. (2009) Definitions and approaches to model quality in model-based software development A review of literature. Information and Software Technology vol 51: Nogueras-Iso J. et al. (2004) Metadata Standard Interoperability: Application in the Geographic Information Domain. Computers, Environment and Urban Systems vol 28: Noy, N. F. and Musen, M. A. (2003) The PROMPT suite: interactive tools for ontology merging and mapping. International Journal of Human-Computer Studies vol 59 (no. 6): Schapke S.-E, Pflug C. and Bögl M. (2012) Multi-models: New potentials for the combined use of planning and controlling information. Mefisto-project, research. Scherer R.-J., Schapke S.-E. (2011) A distributed multi-model-based Management Information System for simulation and decision-making on construction projects. Advanced Engineering Informatics vol 25: De Nier R. (2013), LandXML-OGC. Available at: ESRI Inc. (2003), Spatial Data Standards and GIS Interoperability. Available at: Marine Metadata Interoperability (2013), Metadata Interoperability What Is It, and Why Is It Important? Available at: Network Centric Operations Industry Consortium (2008) Studio Roosegaarde (2011) 179

12 180