Improving Standard Compliance Through Linked Data
A case of product life-cycle management in maritime sector
SUBMITTED IN PARTIAL FULLFILLMENT FOR THE DEGREE OF MASTER
OF SCIENCE
Giedrius Astafjevas
12409324
M
ASTER
I
NFORMATION
S
TUDIES
Information Systems
F
ACULTY OF
S
CIENCE
U
NIVERSITY OF
A
MSTERDAM
3 July, 2019
1st Examiner 2nd Examiner Company Supervisor
Dr.Frank Nack Dr. Tobias Kuhn MSc. Michiel De Vries Faculty of Science Faculty of Science Semmtech B.V.
ABSTRACT
The study investigates how can standard compliance done throughout a product’s life-cycle benefit from the use of linked data. More specifically, the paper examines possibilities of applying linked data in maritime standards. Current standard compliance work in ship-building uses multiple information sources to find and retrieve specifications. This is partly a result of numerous standard issuing bodies (classification so-cieties) that deliver the standards to ship-building companies. Furthermore, the existing compliance methods and systems in place are often identified as lengthy and predominantly in-efficient. This further complicates the standard compliance process. As a potential solution, implementing linked data principles in maritime standards is proposed for enhancing current processes and make standard compliance more user-friendly.
Keywords
Linked data, standard compliance, product life-cycle
management, information management, semantic technology, maritime.
1 Introduction
Efficient information management plays an important role in various aspects of product life-cycle management (PLM) [12]. Throughout the product life-cycle, numerous processes take place that demand for different pieces of information. This information can come from multiple sources provided by different parties, complicating the overall PLM practice [3]. It can often be difficult for companies to clearly define and structure their management processes, which consequentially makes information extraction and use a complex practice [3]. As the implementation of PLM strategies is present in any industry that develops a product, this research takes maritime sector as its main focus. More specifically, ship-building is taken as the main case-study where the use of standards is explored. A vessel’s life cycle goes through three phases of PLM: beginning of life, middle of life, and the end of life. Accordingly, numerous information sources must be consulted to successfully build, maintain, and retire the vessel. The majority of information comes from requirements and specifications outlined in maritime standards, which are commonly referred to as rules in maritime industry. The two terms are used interchangeably in this paper. These standards are the main documents used by engineers throughout
a vessel’s life-cycle. Yet, retrieving and complying to
specifications in standards is often perceived to be inefficient and time consuming work. This is the central problem that is dealt with in this research. It will be argued that this process can be improved by using linked data.
The research takes the use of standards and specifica-tion compliance in ship-building as the primary study objects. It must be noted that only information retrieval part of the compliance process is observed. This approach is taken since information retrieval is the very first step in compliance process, followed by application and verification of specifications. Consequentially, the central goal of the research is to investigate how can the application of linked
data improve standard use and specification compliance in PLM in maritime industry. To successfully answer the central research question, the following sub-question ought to be dealt with first: (1) what are the present issues with the current standard compliance process in maritime sector, and (2) what are the existing practices for applying linked data in standards. This will be done through gathering qualitative and quantitative data, as well as utilising existing software used for linked data application.
There are three methods used in this research:
con-ducting qualitative interviews, collecting survey data, and examining the current application of linked data in practice. The study begins with an overview of scholarly work on PLM and linked data. Afterwards, findings from interview and survey data are synthesised. Lastly, current practices of using linked data in standards are exemplified, followed by the implementation of linked data in maritime standards. 2 Related Work
The following sections present related scholarly work on two relevant topics explored in this research: product life cycle management (PLM) and linked data. While the PLM is consid-erably well studied in the academic literature, a little attention has been given for exploring benefits of linked data in PLM. 2.1 Product Life Cycle Management
The process of product life cycle management is a widely researched domain in academic literature. PLM emerged as a result of a continuously increasing product complexity, higher customer needs, and advancements in manufacturing, project management, and concurrent engineering (CE) techniques [13]. Additionally, companies have to successfully manage themselves in: (1) improving internal and external efficiency (reduce non-relevant costs) and (2) fostering product and pro-cess innovations. Consequentially, a product-driven approach has become another motivation for deploying a well-rounded PLM strategy [14].
From the conception of a product until its retirement (abandon-ment), facilitation and management of information throughout different stages play the key role [13]. Generally, three main phases of PLM are acknowledged: (1) beginning of life (BOL) - product design and manufacturing; (2) the middle of life (MOL) - distribution and maintenance stage; (3) the end of life (EOL) - product retirement and recycle [6]. The process fol-lows chronological order starting with BOL and ending with EOL. In each phase, different information resources are used alongside diverse management techniques. For instance, the BOL stage is based on information used in design and manu-facturing decision making, where information systems such as CAD (Computer Aided Design) and PDM (Product Data Man-agement), as well as knowledge management systems (KM) are widely used to assure a complete flow of information [6]. Information in the whole PLM process transcends the borders of a product itself. It encompasses full intellectual capital of resources used and processes in place. This information is collected in various formats: CAD graphs/drawings, schemes, technical reports, structured and unstructured data, and others [8].
As there are numerous sub-stages and sub-processes occurring in the three stages, some research suggests that PLM process can be too complex to manage for some organisations. This is often attributed to the complexities of sub-processes that occur simultaneously and poor ability to gather and manage information [4]. Therefore, a successful extraction, management, and use of information and knowledge play an important role in efficient PLM [13][4] [14].
Having stated that sound information management is crucial for PLM, some research has been done on the issues with current information management techniques. Multiple information management systems are used in PLM (CAD, CAM, PDM, etc.) to ensure traceability, usability and re-usability of information [9]. Different systems are often used by different parties involved in PLM - clients, producers, suppliers and others, which can lead to information interoperability and modularity issues between different
systems. Therefore, integration of different information
structures that come from different parties present notable difficulties in information management processes. Having different information structures used by different party systems automatically leads to a lower efficiency and accuracy of the whole PLM process [9].
Problems with interoperability and modularity directly relate to the information modelling domain. While some research has been done on the use of XML schema to facilitate information sharing in PLM, a little attention has been given to information model/structure and its semantics[8]. Different products and operations demand for diverse information to be managed successfully. Hence, providing structured information and applying semantics can bring notable benefits. Designing information structure which allows to use metadata to describe contents of the life-cycle, enables user to know which information is needed for each stage of PLM process. Additionally, object characteristics which are brought forward through semantics allow to classify them into different categories and types [8]. As contents of the data are known, classified entities can be further linked together which makes the data model more valuable, interconnected, and easier to use [8].
2.2 Linked data
2.2.1 Linked data principles
Linked data is an approach introduced by Tim Berners-Lee used for publishing and connecting structured data on the web. A set of principles were developed on how to structure and publish linked data alongside standardized Web technologies [2]. The four guiding linked data principles are: (1) using Uniform Resource Identifiers (URIs) for things; (2) URIs should be deferenceable; (3) deferenceable data should lead to discovery of other useful information through URIs; and (4) include URIs in other datasets for further data discovery [2, 15, 6]. Once data is published using these four principles, it can be easily queried with SQL-similar languages, such as SPARQL and accessed by different applications [15]. Addi-tionally, resource-description framework (RDF) is embodied as the main graph-based syntax for structuring linked data [6]. RDF uses a graphic-centered method to make "statements"
about entities using entity-attribute-value triples. Information-linking method is used to "attach" information about resource properties belonging to a specific entity, which creates RDF graph-structures that are retrieved by SPARQL queries [1].
2.2.2 Linked data frameworks
Two frameworks for publishing linked data are proposed. First, Daniel Garijo and Yolanda Gil [7] propose a framework for publishing workflows, where locally developed Wings
applicator is used throughout the process. The generic
sequence of steps found in the framework is as follows: information modelling and conversion, data publication, sharing, and reusing. The framework captures import aspects of linked data. After modelling, workflows are published as RDF triples. At the core of the process, an RDF triple store serves as the database (on a cloud) where RDF workflows are stored. After publication, users can query data through external applications or import the workflows to their local Wings installation. The data querying is done using SPARQL protocol which is supported by the Wings tool-set but can also be queried using any other supporting application [7]. The second framework is applied in the context of emergency management systems and embodies linked data in its informa-tion exchange module [5]. The framework is similar to the one discussed by Garijo and Yolanda. Four steps are executed in the process: exposing data by a collaborative interface, publishing current contextual data as RDF triples, consum-ing and explorconsum-ing combined knowledge as linked data, and providing semantic support. These processes are facilitate by a six-layered structure (four mandatory and two additional layers): data layer, integration layer, application layer, and user layer, with semantic and linked data processing layers being optional to the framework. Each layer is responsible for specific activity that assures the functionality of the upper layer [5].
2.2.3 Linked data in PLM
Only a small number of researches have been conducted on using linked data principles in information management prac-tices in PLM. The most important findings come from a study
carried out by Khaderi et al. [6] who introduce a four-step
process for using linked data throughout different PLM stages. The first step - specification, is used to establish a common knowledge domain that can be shared and re-used in product development. The second step - modelling, creates a semantic model and designs ontology used in the specific product de-velopment context. The third stage deals with generation and publication of data sets, where RDF generated statements are published as linked data on the web. Lastly, exploitation of linked data takes place. Information from different life-cycle stages published as linked data is imported and used by the user. In the end, the research concludes with two substantial findings. First, scattered data throughout PLM is connected and made interoperable to different systems through linked data. This allows to perform automated data extraction and analysis, exchange knowledge, and discover patterns in a do-main. Second, use of ontologies and semantics result in a more uniform knowledge use and re-use that can greatly support decision making in different stages of PLM [6].
3 Research Methodology
Three methods are used in this research to investigate linked data application in maritime standards: interview data collec-tion, survey data colleccollec-tion, and investigating current linked data application in standards. Thus, the research relies on a mixed approach using both qualitative and quantitative meth-ods. The mixed method has been chosen since one method is complimentary to the findings of the other. While qualitative interviews provide valuable insights on the current standard use and compliance process, quantitative data (survey tool) captures more opinions on the matter. In the context of this re-search, survey data is held as complimentary to the qualitative data.
For the interviews, two expert engineers - one with eight and other with twenty years of experience in standard compliance were approached from two major dutch ship-building compa-nies - Damen and Feadship. A total of four open questions were asked (see Appendix B). To conduct a survey, a group of eight engineers, whose experienced varied between five and twenty years, from different ship-building companies were asked to participate in the survey. The questionnaire was build using Google Forms online tool and consisted of 20 close-ended questions (see Appendix A). LinkedIn platform was used as the primary distribution method to collect responses. 4 Results
The following chapter presents findings on three instances. First, findings from qualitative interview data are discussed, followed by the results from quantitative survey data. After-wards, two process frameworks on standard compliance are introduced: (1) the current standard compliance process frame-work in maritime industry and (2) the proposed linked data framework for standard compliance.
4.1 Qualitative interview data: current compliance pro-cess
The related work section has shown that scholars often identify information management as one of the main criteria
for successful product management[4], [13], [14]. Two
interviews were conducted with two engineers from Damen and Feadship - the two largest ship-building companies in the Netherlands. Each interview lasted for approximately one hour and allowed to better understand the current standard use and compliance process, as well as the commonly experienced issues. Findings can be briefly summarised as follows. It was found that standards, commonly referred to as rules in ship-building, come from different classification
societies. Classification societies are the main standard
publishers that deliver standards to engineers. Considering the number of societies, two notable issues in the current standard compliance work done by engineers are found: scattered information and manually performed compliance tasks. Additionally, these two issues are further facilitated by the absence of effective software capable of improving standard compliance. The following sections discuss the interview findings in a greater detail.
The largest and most relevant societies are International Maritime Organisation (IMO), Bureau Veritas (B.V), and
Lloyds Register1. Additionally, the flag state - a country for
which the ship is being built, can deliver an additional set of specifications that must be complied to. As a result, the multiple bodies that provide specifications require engineers to search different rules until all needed specifications are retrieved. This entails that cross-referencing is a common practice in the current composition of rules. For example, when designing a hull of the ship, engineer can use a rule published by B.V. but at some instance be referenced to a different standard, where the needed specification is outlined. Such structure can lead to an extensive list of documents where specifications must be retrieved from
different standards. Furthermore, it was also found that
specifications are sometimes scattered throughout the same document. This discloses the first issue with the existing compliance process - scattered information.
Scattered information would not present as much of an issue if the process is automated, which would allow to locate and retrieve specifications automatically via a system. However, current methods are predominantly based
on manual work. The most common standard format is
a Portable Document Format (PDF). Rules mostly have a uniform structure: index, system topics (general categories), and object/activity/process paragraphs assigned per topic, which provide more detailed information on a given object. While this structure gives some level of guidance to the user, it is a manual process to search the file (or multiple files) for a desired specification. The whole sequence of steps that starts with receiving a PDF file, finding and extracting the needed specification, and applying it in the product management process, has been described by engineers as predominantly
inefficient due to lack of automatized functions. Hence,
this disclosed the second problem with the current standard compliance framework - manual-work based tasks.
The last finding entails that some level of database-centred information structure is used. Some classification
societies like B.V and Lloyds have a "Rule Finder"23tool.
The tool permits to search the database created by the classifi-cation society, where published rules are stored. While this system has not been investigated and tested in this study, the interviews disclosed that there are a few notable issues with the tool. First, the database only contains rules published by the society that provides the rule finding service. Other rules that come from different societies are not stored on the same database and hence, cannot be searched. Second, the tool only permits to manually search for a specific keyword like a "hull" or an "engine" without filtering options. Consequentially, ambiguous results are returned. The third issue with the rule finder is its inability to execute specific queries that could automatically retrieve the requested information from the database. Instead, the user must manually search the database using the keyword finder. Whilst the concept behind such tool is worthy to explore, the inner structure and available query 1https://puc.overheid.nl/nsi/doc/PUC
130114/ 2Lloyds:https://www.lr.org/en/rulefinder/
3
options do not provide the necessary support to the user. The rule finder tool is not explicitly analysed in this research but future work could investigate the potentials of extending the software functionalities for processing linked data.
Based on the results from the interview data, scattered information and manual-work based tasks result in inefficien-cies and lengthy procedures that delay product development. Additionally, the rule finder - a useful but not a fully empowering software tool, further complicates compliance work. Findings from the interview data allow to create a framework that captures the current standard compliance process. This framework is shown in Figure 1.
Figure 1. Current standard compliance framework
As shown in figure 1, the user utilises four standards from four different classification societies in his work. Lloyd’s and B.V have the rule finder software in place where specifications in standards can be searched manually. IMO and Flag State standards do not provide a user with such software. To fully comply with specifications, the user is required to check differ-ent PDF files until all the needed information is retrieved and can be used in the later compliance stages [11]. Considering the issues with the current standard compliance practices in maritime sector, it is suggested to improve the current frame-work. To do so, a new approach facilitating linked data is presented in section 5 after disclosing a list of requirements. The following section will discuss the results from survey data that compliment the interview data findings.
4.2 Survey Data
The survey data was used to capture more opinions of engi-neers on the current standard compliance process. The total of 8 ship-building engineers from different companies were selected and asked to complete the survey. Paragraphs below
summarise the most important survey findings on three as-pects: the use of rules, the structure of rules, and specification compliance tasks.
4.2.1 Use of rules
This section covered the most important rule usability criteria. When asked to rate the importance of rules on a scale from one (the worst/lowest) to five (the best/highest), 38% of engineers gave a rating of four. The remaining 62% rated the importance with a five which indicates that the use of rules is of significant importance. On the same scale, when asked about how easy it is to use the rules, the majority (63%) gave an average rating of a three, whilst 25% rated with a four and 12% with a two. This shows that most of engineers experience some level of hardship when working with rules.
The two major usability aspects - time efficiency and the amount of manual work in the process, received a compa-rably negative evaluation. A large portion - 63% consider working with rules to be a time-consuming process, whilst 75% perceive rule compliance to be predominantly based on manual work. Lastly, data shows that more than half of the respondents (62%) often use different rules (PDF files) to get all the necessary information.
4.2.2 Rule structure
Here, collected data show user satisfaction with rule structure. On a scale from one to five, 38% rated their satisfaction with a rule structure with a three. The remainder 25% chose four and five, with 37% of participants being unsatisfied and giving ratings of one and two. As shown by the interview data, infor-mation is often scattered making it difficult to locate. This finding aligns with survey data. Three quarters of respon-dents found information to be often scattered in rules, with the remainder 25% only sometimes encountering information scarcity issue.
When inquired whether better structure and organisation of rules would make them easier to use, the answers were nearly uniform. 88% agreed that better rule structure would make them easier to use in practice. The remaining 12% did not know whether improved structure would help.
The last question dealt with applying information models in rules. A large portion of respondents (75%) believed that information models could improve usability of rules. The remainder did not know whether information models would help to improve the compliance process.
4.2.3 Specification compliance
In this part, data was collected on two instances: how well-written are specifications; and how easy it is to comply
to specifications within rules. First, SMART (specific,
measurable, acceptable, realistic, time-bound) criteria [10] in specifications was evaluated. On a scale from one to five, half of the respondents rated quality of specifications with a three. This demonstrates that current descriptions are merely acceptable. Furthermore, 25% gave a five star rating whereas 12% evaluated specification descriptions with only two stars. These results show that SMART criteria is not fully met. To continue, a large number of engineers (50%) perceived connectivity / linking between specifications in different information sources to be rather poor. The second biggest
group (38%) rated connectivity with only two stars. Similarly, when asked about how easy it is to search/locate specifications in rules, half rated the process with only two stars. The remaining 38% attributed three stars and only 12% were satisfied with the search procedure.
In the end, the participants were asked to select (multiple-choice available) which problems do they encounter most often when working with specifications. This question has helped to establish which problems are the most prevalent. The enumeration below lists the most commonly identified issues.
1. Finding and checking specifications takes a significant amount of time - encountered by 88%.
2. Majority of specification compliance work has to be done manually - encountered by 75%.
3. Information is scattered: different sections or different rules have to be used to find information on one specifications -encountered by 50%.
After completing the analysis, it can be stated that the survey results align with the interview data findings presented in sec-tion 4.1. Survey results reconfirm the two biggest issues found with the current use of standards and specification compliance: scattered information and manual compliance work. Having reconfirmed the initial interview findings with a broader sur-vey data, the section below distills a list of requirement for improving the existing compliance process.
4.3 Requirements for improving compliance
Interview and survey data findings permit to draw a set of requirement for improving the standard compliance process in maritime-sector. These requirements are based on the needs of engineers and focus on improving information management throughout PLM in ship-building. As will be shown throughout this research, these requirements could potentially be met by linked data.
The first requirement deals with scattered information problem. It is not feasible to assemble all specifications from different standards and publish them as a single file, since they are issued by different societies. Thus, specifications will remain scattered across standards. Consequentially, the first requirement entails that specifications across different standards are better linked and can be retrieved more easily. Creating links between objects and adding semantics can aid engineers in knowing which specifications belong to which object, as well as what are the relationships between different specifications.
The second requirement addresses the manual work is-sue. A large number of survey respondents (75%) indicated that most of the compliance work has to be done manually by engineers. Presently, maritime standards are published in non-machine readable data formats (mostly PDF) and thus, cannot be processed automatically. This permits to draw a second requirement to transform current data formats into machine-readable structures.
The last requirement is derived from the two above mentioned conditions. Therefore, it is not a direct requirement and will be treated as an additional criteria throughout this research. To improve current compliance practices a system
capable of processing machine-readable data formatsshould
be developed / introduced. In this research, Laces suite is used as facilitating software for linked data application.
4.4 Proposed Framework
Reflecting on the issues found in the present standard com-pliance process (sec 4.1 and 4.2) and the list of requirements (sec 4.3), it is suggested to introduce linked data structures into the existing standard compliance framework shown in Figure 1. Parts of the two frameworks described in related work chapter (sec 2.2.2) are adaptable for the new compliance process. The proposed framework is similar to the one intro-duced by Kadiri et al. 2013. Their research suggested four activities for introducing linked data in PLM: specification, generation, publication, and exploitation/extraction [6]. This is translated into a proposed standard compliance framework that incorporates linked data and is shown in Figure 2.
Figure 2. Proposed linked data framework
Four chronological steps occur in the compliance framework in Figure 2. First, information in a standard (PDF file) is captured and modeled into Object Type Libraries (OTLs) and Specification Libraries (SLs). This is a novel step specifically introduced in this framework. Once the OTLs and SLs are created, they are transformed into RDF triples using specific vocabularies (ontologies). In the third stage, RDF triples from different standards are published on a linked data platform. The platform is initially a large triple store with millions of triples from different information sources. In the last step, applications are given access to the platform and are able to execute queries using SPARQL protocol. The proposed framework from Figure 2 is applied in practice in section 5.1. 5 Linked Data in Standards
The previous chapter disclosed the most important data collec-tion findings and introduced the proposed linked data frame-work (Figure 2). The following sections in this chapter will be used to exemplify implementation of such framework. First,
linked data application process is demonstrated (sec 5.1), where the results from an already implemented project that uses linked data are shown. This creates the basis for using linked data in standards. Afterwards, the same method is ap-plied for implementing linked data in maritime standards (5.2). The chapter will conclude with validation and verification of the final results (sec 5.3).
5.1 Applying linked data: Laces software
The proposed compliance framework shown in Figure 2 in sec-tion 4.4, presents a four-step structure for modelling, applying, sharing, and accessing linked data. Laces software suite
devel-oped by Semmtech4adopts a similar four-step process and has
therefore been chosen to implement the proposed framework. Figure 3 portrays the steps taken when using Laces software.
Figure 3. linked data Publication Process
There are four chronological steps in the process: (1) construc-tion of OTLs and SLs; (2) structuring data as RDF, (3) pub-lication of RDF triples on the platform; and (4) user-enabled SPARQL queries. Below, each step is analysed in a greater detail. To exemplify the process, a standard from construction industry was used since it has already been implemented using Laces software. The standard’s primary information source is Annex 14 to the Convention on International Civil Aviation (later referred to as Annex 14), issued by International Civil Aviation Organization (ICAO) in a PDF format.
5.1.1 Step 1: Information modelling
The very first stage of the process models information through Object Type Libraries (OTLs). Laces Library Manager is the primary tool used for constructing OTLs. It must be noted that this is a manual process, requiring user to manually built the information model of a standard. OTLs are based on a "parent-child" logic, where parent categories have children sub-categories. Next to standard OTLs, the software can be used to create additional Specification Libraries (SLs). SLs share the same composition to OTLs and categorize specifi-cations found in a document. Figure 4 shows an example of OTL and SL modelled from Annex 14.
As shown in Figure 4, OTLs and SLs present basic asset infor-mation modeling from Annex 14. OTLs create a single and standardized concept-base of different object categories and types, which are described using a standard/uniform language. This makes information exchange more uniform and accurate
4Laces software tools: https://semmtech.com/products
Figure 4. Object and specification type libraries in Laces Library Man-ager
in its use and re-use. These standardized object types can be easily managed and shared across different stages of a product life-cycle. Standardized information exchange across differ-ent processes in PLM can lead to an improved managemdiffer-ent of information[12]. The same logic applies for SLs where specifications can be standardized and structured according to different categories and assigned to specific objects. As shown in the figure above, three types of conceptual specifications are made: notes, recommendations, and specifications. Speci-fications 2.2.3 and 2.3.2 are highlighted as they are explicitly used when illustrating the next steps of the process.
5.1.2 Step 2: Linked data application
The second part of the process is transforming object and specification type libraries into RDF triples. This is a fully automated intermediary step occurring between Laces Library Manager and Laces linked data platform (LDP). One or a combination of vocabularies can be used in the triple structure,
such as OWL5, Dublin Core6, SKOS7, and others. The choice
for a specific vocabulary(-ies) is predominantly determined by the agreement between the information provider and the end user. The agreement establishes which vocabularies are the most useful and widely applicable in the given context. Thus, the scale and scope of information use, such as the targeted users and the intended domain, play an important role.
5OWL ontology: https://www.w3.org/OWL/
6Dublin core ontology:
http://www.dublincore.org/specifications/dublin-core/dcmi-terms/
7SKOS ontology:
RDF represents OTL and SL libraries in a machine-readable data format that can be easily read by different applications. An important aspect of linked data is assignment
of Unique Resource Identifiers (URIs). Each published
object/subject or a specification is given URI which permits to easily refer to any given object/specification on the web [1]. Figure 5 shows specification 2.2.3 from Figure 4 transformed into RDF triple.
Figure 5. Structuring data into RDF
The machine-readable presentation of the triple in Figure 5 uses CSPEC vocabulary. The CSPEC is based on SKOS ontology and introduces more detailed vocabulary for this project. This vocabulary was extended by adding a part of Dublin Core (dcterms) ontology that identifies the "type" of an entity. There are six prefixes composing the triple: name,
subject, type, specifies, description, and source(highlighted in
bold in Figure 4). While some prefixes use plain text in their structure, such as description and name, others contain a URI to another object/activity/entity, etc. URIs act as identifiers of an object/specification/entity, which are retrieved and can then be displayed as a plain text, as shown in Figure 5. A salient feature of this triple is the prefix "heeftBron" identifying the source of information. This allows application, and later the user, to know from which specific source the data has been retrieved (in this case the URI refers to Annex 14 document). The human-interpretable format is given as an example of what a person would understand from the given RDF triple. Transforming information into machine-readable data can po-tentially increase the overall efficiency of information use and re-use, as well as the accuracy of this process. The next step will present how these triples are published and stored in a large triple database on LDP.
5.1.3 Step 3: data publication on the platform
After data has been transformed into RDF triples it is published on LDP. LDP is initially an extended version of a standard triple store with some added features, such as version control and authentication mechanisms. The platform stores multiple data sets of RDF triples from different information sources. Once granted access, these triples can be easily retrieved by dif-ferent applications through add-on enabled SPARQL queries. Each URI on the platform presents a unique entity. When publishing specifications, three types of entities can exist on
the platform: specification, activity, and object. An important feature of the platform is enabling access for different applica-tions to retrieve these triples. The platform would partly lose its value if the triples could only be stored but not queried and retrieved by the intended users. Consequentially, accessibility criteria met through HTTP permits communication and data exchange between the platform and different user applications. To retrieve specific information pieces from the rich triple ecosystem, SPARQL queries are executed by add-on software. The last step explains the query process in a greater detail.
5.1.4 Step 4:Querying linked data on the platform
In the last stage of the process, once linked data has been published on the platform, the user applications could retrieve the data from LDP. Some of the existing applications tend to not have SPARQL querying mechanisms built-in. It is possible to extend such software with the functionality, of course, or to employ an add-on - a small communication module that permits querying. The add-on is able to communicate with both - user application and LDP given the HTTP protocol. Depending on the user application configurations, add-ons are designed accordingly to facilitate data exchange between the two systems and thus, improve data interoperability. The add-on enabled data flow is presented in Figure 6.
Figure 6. Add-on facilitated data exchange
This process has four chronologically occurring steps. First, the user pre-selects the information he wants to retrieve from the LDP. This is done on the add-on configured for the user software. An example of user add-on interface is given in Figure 7.
The user interface given in Figure 7 allows the user to select information he wants to retrieved from the platform. In the list of specifications, all specifications created from Annex 14 are given. Once the user selects which specification triples to import from LDP, in this case 2.2.3 and 2.3.2, the add-on sends a SPARQL query to retrieve the matching triples. Given the information request, the add-on executes the query shown in Listing 1.
Listing 1. SPARQL query for retrieving specifications from the platform SELECT ? S u b j e c t ?Name ? D e s c r i p t i o n ? S o u r c e ? Type WHERE{ ? s p e c a c s p e c : S p e c i f i c a t i e . ? s p e c c s p e c : s p e c i f i c e e r t ? S u b j e c t . ? s p e c c s p e c : naam ?Name . ? s p e c c s p e c : b e s c h r i j v i n g ? D e s c r i p t i o n . ? s p e c c d o c : h e e f t B r o n ? S o u r c e . ? s p e c d c t e r m s : t y p e ? Type . }
This query retrieves all the requested information available about selected specifications 2.3.2 and 2.2.3. Once the data has been queried, the information is sent back to the add-on software. Before importing data back to user application, data retrieved through SPARQL query is converted into a readable data format (such as XML) for the user application. The whole process is fully automated and the only input done by the user is selecting the needed information to be queried by the add-on software.
5.1.5 Displaying retrieved results
As the data is returned to the user application, the visual display and structure of retrieved data depends on the application’s interface. Overall, the user will be given an
overview of the requested information. Looking at the
SPARQL query example in Listing 1, the user will receive detailed information on the following specification details: full name, type, source, description (text) of the specification, as well as the subject to which the specification relates to. As noted, the representation of the results predominantly depends on user application’s interface. The returned results in Relatics user interface are shown in Section 5.2, where specifications from a maritime standard are retrieved and shown.
Overall, using the proposed linked data framework embodied by Laces software can be a novel method for publishing maritime standards. As discussed by scholars [2, 4, 7], one of the greatest benefits of linked data is connecting different pieces of information and establishing relationships between entities, with the help of URIs. The proposed linked data framework brings the following benefits:
1. The four steps facilitate an efficient exchange of data be-tween the information provider and the user.
2. OTLs and SLs classify asset information that can be traced and linked throughout different projects. This increases the accuracy and uniformality of data.
3. Information published as linked data is machine-readable and can be used by different applications. SPARQL per-mits an easy and effective way to query data, where linked specifications can be retrieved on a specific object. Due to assignment of URIs and linking between entities, informa-tion about specific object can be retrieved from multiple information sources, for instance different standards. This could be further extended into verification stages of a prod-uct.
The following chapter explores how linked data framework, originally showed in Figure 2 and exemplified using Laces software, can be applied in maritime standards.
5.2 Implementing linked data in maritime standards The findings from interview and survey data presented in sections 4.1 and 4.2, disclosed two key issues with the current compliance process: scattered information and manual compliance work. Collected knowledge permitted to draw three requirements (sec 4.3) for improving the process: linking specifications (connecting data), automatising compliance work (machine-readable data), and developing/introducing a facilitating software. The following sections will implemented linked data framework shown in Figure 2 to maritime standards. First, linked data application using Laces software
will be discussed and the returned results in Relatics8interface
will be presented. Relatics software has been chosen as it is a popular information management tool across different industries. Afterwards, validation and verification of the results will be done.
The Laces software suite was examined in section 5.1 to exemplify how can information in a PDF standard be modelled, structured, published, and retrieved as linked data. Same method can be applied for maritime standards given the fact that they are published as PDF files. For using linked data in maritime standards, the four steps detailed in section 5.1 were followed and a standard published by Bureau Veritas (B.V) on Rules for the Classification and the Certification of
Yachtswas used. The first three stages are discussed briefly as
their processes are identical to the ones presented in previous section. The majority of attention is given to the analysis of the retrieved results in stage four.
First, information in a standard is modelled as object and
specification libraries using Laces library manager. The
result is identical to the library example given in Figure 4, differing only in the contents of found specifications. A part of text transformed into libraries can be found in Appendix D. Specification numbered as 3.1.5 was used in the next steps. The second and the third steps deal with making data machine-readable and publishing data on the linked data platform. As shown in section 5.1.2, specification library is transformed into RDF structure, which is a fully automated step occurring between the Laces library manager and the 8Relatics software: https://www.relatics.com
platform. Once the data is structured as RDF triples, it is made available on the platform, as discussed in section 5.1.3, and is ready to be queried.
In the last part of the process, step 4, triples can be queried and imported by user applications. The same query as shown Listing 1 was used to retrieve the desired data from the platform. As noted before, not all applications have querying functionality and thus, add-on configurable software can be added as an extension. When applying linked data to a maritime standard, Relatics software - a web-based project and information management tool, was used to display the queried specifications. Figure 8 shows user importing specification 3.1.5 and receiving results on Relatics interface.
Figure 8. Imported specification in Relatics interface The results displayed in Figure 8 are based on the executed SPARQL query, which had a similar composition to the query shown in Listing 1. The query retrieved data about specifica-tion’s title, status, version, description (text), source URI, and physical object. The displayed results are cardinally differ-ent in comparison to what an engineer currdiffer-ently sees in PDF standards. Instead of having a plain-text file where specifica-tions have to be manually searched and extracted, they can now be automatically retrieved and displayed by a software. As discussed before, one of the biggest problems with the current composition of standards is manual compliance work, i.e. looking for scattered specifications. Given the machine-readable RDF format, specifications published as linked data can be easily queried, imported, and read by different appli-cations, such as Relatics. Machine-readable data eliminates the need to manually search for specifications and the results can be displayed in a logical manner in user application.
Addi-tionally, given linked data principles, specifications are linked
to the corresponding physical object, in this case a collision bulkhead. These linking principles can be taken a step further by linking and retrieving all specifications that are applicable to this object. This is shown in Figure 9.
In contrast to Figure 8 where the main query object is a specification itself, Figure 9 depicts retrieval of all speci-fications from a specific object - a collision bulkhead. A graph-based information structure about the object is retrieved and three specifications are found. Considering the fact that
Figure 9. Retrieving all specifications applicable to one object
RDF structures can have unlimited links, more specifications from different sources can be found, as well as other related objects. From the executed query it is known that the object has a name: collision bulkhead, object URI, and compliant specifications 3.1.4, 3.1.1, and 3.1.5. Even though in this example the displayed specifications are from the same source (B.V rules classification for Yachts), they can (and usually do) derive from different standards. Linking different pieces of information and retrieving them as a single unit/graph, arguably makes the compliance process more organised. A major problem discovered with the current use of standards was scattered information. Linking between different informa-tion pieces done using RDF permits to retrieve informainforma-tion on a specific entity despite its location/source. Hence, such approach potentially solves the scattered information problem in maritime standards.
Overall, the shown results in Figures 8 and 9, represent a different spectrum on how specifications in maritime standards can be located, retrieved, and imported into a local user application. Using linked data in standards is a relatively novel but a promising approach. The two notable advantages of using linked data in maritime standards, as already illustrated with the heliport example in section 4.5, are easily retrievable, processable, and linked data. The section below validates these findings.
5.3 Validation and Verification
To prove usefulness and validity of the proposed linked data method, validation and verification of the results were carried out. The purpose of validation is to validate whether the research findings (using linked data in maritime standards) meet the initial requirements. Verification, on the other hand, is done to make sure that the implementation of the results is error-free and executed in accordance with the established practices.
5.3.1 Validation
To validate whether the research results meet the requirements outlined in section 4.3, three interviews were held with relevant experts: two ship-building engineers and chief executive officer of Semmtech. Inquiries were made on three key aspects:(1) overall added-value of linked data approach; (2) evaluation of compliance process efficiency; and (3) evaluation of information connectivity.
First, all three interviewees expressed a uniform opinion that linked data approach would be a valuable addition to
the current standard compliance practices. Comparisons
were made between the current and the proposed process frameworks. Advantages of linked data, namely machine processable data and linking between different pieces of information, resulted in experts acknowledging the overall added-value of linked data.
Second, both engineers agreed that the process efficiency, in terms of time spent working with standards, would improve if parts of the compliance work, such as location and retrieval of specifications, are automated. More specifically, machine readable data was the key characteristic leading to a more efficient standard use due to provided task automation. The following remark was made by one of the engineers:
"The data is further scaled to machine readable level and query method helps to reduce the time spent looking for specifications. To me, machine readable data is useful because it could replace manual searching in PDF files".
Third, connecting information in standards would solve infor-mation scarcity issue. The experts gave positive feedback on establishing links between different entities. Linking objects to specifications using linked data was seen as a valuable at-tribute of linked data by all three interviewees. The following is an example of a comment made by one of the interviewees: "I consider linking specifications to assigned objects as a very useful aspect. Knowing the type of information and where it comes and to which entities it relates, can save a lot of time when working with standards".
Based on the comments by the three experts, findings were successfully validated on all three instances - added-value, process efficiency, and information connectivity. Positive feed-back given by both engineers and the CEO of Semmtech per-mit to conclude that linked data can improve current standard compliance practices in maritime sector. From the interviews conducted, it could be further stated that linked data can po-tentially solve currently present issues in standard compliance work.
5.3.2 Verification
To verify whether linked data solution is implemented correctly, experts at Semmtech were consulted throughout the whole implementation process. Since Laces software suite was used to demonstrate linked data applicability in standards, verification was done on the following aspects: logic and quality of SLs, publication process on the platform, configuration of the add-on, and quality of the retrieved results. After the end of each process, the result was compared to the results of an already implemented project shown in 5.1. In case of notable disparities, improvements were made to match the existing project. Using linked data in standards is a chronological process consisting of four steps: modelling, transformation, publication, and querying. Hence, an error-free step-by-step implementation of each stage will automatically lead to final results displayed in Figures 8 and 9.
Overall, each step of linked data application to maritime stan-dards was verified throughout the whole process. Nevertheless, to have more opinions on the methods used in this research, it
canbesuggestedtofurtherverifythefindingswithexternal
companies/experts,aswellasclassificationsocieties.
6 Discussion
Throughout this research, application of linked data in
standardcompliance hasbeenexplored. Morespecifically,
linkeddatahasbeenintroducedintothestandardsofmaritime
industry.Thediscussionchapterwillconsiderthefindingson
threeinstances:reflectionontheresults,researchlimitations,
andproposeddirectionsforfuturework.
Reflectionontheresults.Beforeproposingtouselinkeddata
toimprovestandardcomplianceinmaritime,three
predomi-nantissueswiththecurrentprocessweredisclosed:scattered
information,manual-workbasedcompliancetasks,andlack
offacilitatingsoftware. Reflectingontheseproblems,three
requirementswereraisedforimprovingtheprocess:linking
specifications in standards, introducing machine-readable
datastructures,andadoptingasoftwarecapableofcapturing
the two above mentioned requirements. As a result, the
researchproposedanovelstandardcomplianceframework
wherelinkeddataisused.Toverifytheapplicabilityofthis
framework, analysis on current practices of using linked
datainconstruction standards(section4.3)wasdone. The
sameapproach wastakenforimplementinglinkeddata in
maritime rules (section 5.2). After considering the final
findings,itcanbestatedthatitisfeasibletouselinkeddata
approach in maritime rules. Furthermore, adding to the
feasibilityaspect,itcanbearguedthattheproposedapproach
meetsthethreerequirementsstatedabove.RDFdatamodel
potentially tackles all the three major issues in maritime
sector and improves the standard compliance process by
linkinginformationandpresentingdatainamachine-readable
format. Though, oneinstanceoftheprocess-information
modelling(showninFigure4)stillhastobedonemanually
bytheengineers.Everydesiredstandardhastobemanually
transformedintoobjecttypeandspecificationlibrariesbefore
publishing it aslinked data. Even though modelling only
needstobedoneonce,itcanbeatime-consumingprocess.
Twopotentialsolutionsareproposedtoavoid/improvethis
step:(1)introducealgorithmsthatarecapableofmodelling
text into concrete information models; (2) use external
companiestooutsourcethispartoftheprocess.
When reflecting on the overall feasibility of linked data
method,ithasonly beenprovenonatechnicallevel. The
study has shown that it can be done using the available
software tools, such as Laces and Relatics. However,
implementationinreal-lifesituationhasnotbeentested. It
remainsunclearwhetherthebureaus/classificationsocieties
wouldbekeenonusinglinkeddatainstandardsasitwould
alterthecurrentbusinesses modelsinplace. Even though
engineers would benefit from such approach, contrasting
views may be received fromthebureaus. Inother words,
whilethisstudyshowsthatlinkeddatamethodcanbeused
instandards, it doesnot permitto concludethat itwill be
adoptedinreality.
Consideringtheoverallscopeofthisresearch,itcanbe argued
thattheresearchdoesnot fullycapturetheentirestandard
compliancecycle(location,retrieval,application,verification).
standard compliance process are considered - locating and retrieving needed data (specifications) from standards. After data is located and retrieved, using specifications in practice and verifying whether their application is done correctly takes place. The research does not investigate linked data application in these two processes. It has been the goal of this study to only investigate the first two steps of the whole process due to time and resource constrains. Thus, the scope of this research was purposefully limited to the location and retrieval of data. In terms of the software adopted for this research - Laces Suite, it is not a condition to use this particular application. Other applications that are capable of publishing and querying linked data can be employed, such as Talis Platform and Pubby, as well as a variety of other semantic web browsers and libraries. The same holds true for the user-end application that receives the SPARQL query results. In this research, specifications from a maritime standard retrieved as linked data (Figure 10 and 11) are displayed using Relatics interface.
However, alternative software tools, such as IBM Doors9or
Outsystems10, could be employed by engineers. The section
below addresses one of the key limitations of this research - investigating user applications that are currently used by
engineers in ship-building.
Limitations. To have a full-rounded interpretation of the
research results, study limitations ought to be discussed. First, limitation on the survey method is outlined. The total number of eight survey participants may be considered as one of the research limitations. Such sample size is not sufficient
to establish statistical relevance / significance, neither that
is the goal of the survey. The purpose of the survey is to gather more opinions on the prevalent issues in the current standard compliance. Having stated that survey data plays a complimentary role next to the interview data, the number of respondents is considered to be sufficient. Nevertheless, it would be desirable to have a larger sample size in future research work.
The second limitation derives from the "facilitating" software used for publishing and querying linked data. Only one tool - Laces software suite, was tested. To determine the most optimal facilitating software, alternative applications that have linked data publication and querying capabilities should be investigated. Ideally, a criteria is developed for evaluating each application and then determining, which system is the most suitable choice for facilitating linked data. This relates to the limitation dealing with the current software used by engineers. The research did not investigate which software tools are currently available when working with standards and it cannot be said for sure that their functionalities cannot be extended to facilitate linked data application.
9IBM Doors:
https://www.ibm.com/nl-en/marketplace/requirements-management
10Outsystems: https://www.outsystems.com/platform/
The last notable limitation of this work is in its original scope. Application of linked data in standards has only been explored from the engineer point of view. Consequentially, classification societies/bureaus are not considered in this study. An important aspect deriving from this limitation is the lack of knowledge about the "Rule Finder" tool, developed and
used by Bureau Veritas and Lloyds. Knowing more details about this tool is necessary for deciding upon the facilitating
software. For instance, facilitating software may not be
needed if the Rule Finder functionalities can be extended, enabling software to read and query RDF structures. Whilst this has not been analyzed in this research, it could be done in future studies.
Directions for future work. Considering the reflection on the re-sults and limitations of this work, there are a few instances on which more extensive research should be done. First, linked data application throughout the full compliance process (in-formation location and retrieval, application, and verification) should be investigated. Future researches could further verify the proposed method outlined in this study and extend the approach to application and verification steps. Second, future studies should employee a larger sample size in their research. While this research can be considered to be an exploratory study, gathering more data from engineers, classification so-cieties, and other relevant stakeholders (such as 3rd party companies) is advisable. Lastly, it is suggested to explore the capacity to integrate linked data with current systems in place, such as the "Rule Finder" tool. This can be done along-side the studies that will predominantly focus on classification societies.
7 Conclusions
The central goal of this study was to investigate how can standard compliance be improved through linked data. To do so, a novel framework that incorporates the use of linked data in maritime standards has been introduced.The analysis has shown that linked data could improve the present process on two instances: making specifications in standards linked and machine-readable. This solves the two major issues found in the current compliance work - scattered information and manual compliance tasks. Furthermore, exploring linked data practices in construction industry has given empirical grounds for this research as it demonstrated technical feasibility to apply linked data in standards. When weighing pros and cons of the proposed linked data approach, it can be concluded that standards in maritime sector can indeed benefit from the application of linked data.
This study aims to contribute to the existing knowl-edge base in both, the maritime industry and academia. A limited number of scholarly work investigates how can applying linked data in standards and specifications (in any industry/sector) be beneficial for its users. It is hoped that this research improves the available knowledge on the subject and lays the foundation for future studies in the fields of standard compliance and information management.
8 Acknowledgments
Advise and supervision by UvA supervisor Frank Nack, Semmtech supervisors Michiel De Vries and Sander Stolk, were of pivotal importance for successfully completing this project.
References
[1] ANTONIOU, G.,ANDVANHARMELEN, F. A semantic
web primer. MIT press, 2004.
[2] BIZER, C. The emerging web of linked data. IEEE intelligent systems 24, 5 (2009), 87–92.
[3] CANTAMESSA, M., MONTAGNA, F.,ANDNEIROTTI,
P. An empirical analysis of the plm implementation effects in the aerospace industry. Computers in industry 63, 3 (2012), 243–251.
[4] DANIELS, M., LANNING, I., GHADIMI, P., HEAVEY,
C., RYAN, A.,ANDSOUTHERN, M. Product lifecycle
management requirements gathering: Industrial pilot cases. IFAC Proceedings Volumes 46, 9 (2013), 1750–1755.
[5] DEFARIACORDEIRO, K., MARINO, T., CAMPOS, M.
L. M.,ANDBORGES, M. R. Use of linked data in the
design of information infrastructure for collaborative emergency management system. In Proceedings of the 2011 15th International Conference on Computer Supported Cooperative Work in Design (CSCWD) (2011), IEEE, pp. 764–771.
[6] ELKADIRI, S., MILICIC, A.,ANDKIRITSIS, D.
Linked data exploration in product life-cycle management. In IFIP International Conference on
Advances in Production Management Systems(2013),
Springer, pp. 460–467.
[7] GARIJO, D.,ANDGIL, Y. A new approach for
publishing workflows: abstractions, standards, and linked data. In Proceedings of the 6th workshop on
Workflows in support of large-scale science(2011),
ACM, pp. 47–56.
[8] JUN, H.-B., KIRITSIS, D.,ANDXIROUCHAKIS, P.
Research issues on closed-loop plm. Computers in industry 58, 8-9 (2007), 855–868.
[9] LEDUIGOU, J., BERNARD, A.,ANDPERRY, N.
Framework for product lifecycle management
integration in small and medium enterprises networks. Computer-Aided Design and Applications 8, 4 (2011), 531–544.
[10] MANNION, M.,ANDKEEPENCE, B. Smart
requirements. ACM SIGSOFT Software Engineering Notes 20, 2 (1995), 42–47.
[11] MUSTAPHA, A., AROGUNDADE, O., VINCENT, O.,
ANDADENIRAN, O. J. Towards a compliance
requirement management for smses: a model and architecture. Information Systems and e-Business Management 16, 1 (2018), 155–185.
[12] O’KEEFFE, AONGHUS, ALSEM, DAAN, CORBALLY,
ROBERT, LANENVAN, R. Information Management for
European Road Infrastructure using Linked Data -Investigating the Requirements. Tech. rep., Conference of European Directors of Roads, 2017.
[13] OLIVEIRA, P. S. G.D., SILVA, D.D., SILVA, L. F.D.,
LOPES, M.D. S.,ANDHELLENO, A. Factors that
influence product life cycle management to develop greener products in the mechanical industry. International Journal of Production Research 54, 15 (2016), 4547–4567.
[14] TAISCH, M., CAMMARINO, B. P.,ANDCASSINA, J.
Life cycle data management: first step towards a new product lifecycle management standard. International Journal of Computer Integrated Manufacturing 24, 12 (2011), 1117–1135.
[15] YU, H. Q., PEDRINACI, C., DIETZE, S.,AND
DOMINGUE, J. Using linked data to annotate and search educational video resources for supporting distance
learning. IEEE Transactions on Learning Technologies 5, 2 (2012), 130–142.
1/5
Survey Research
Dear participant,Thank you for taking your time to complete this survey. The survey consists of 20 closed-ended questions and will take no longer than five minutes to complete.The points below provide the most important information about this survey research.
1. Survey's purpose.
The survey is used to capture engineer experience when using different rules/standards (from B.V, Lloyds, IMO, others) in specification compliance process in maritime sector. More specifically, the data collected will help to understand the daily problems encountered by engineers when working with rules. 2. Anonymity.
In this survey, it is not requested to provide any personally identifying information (PII), such as names, phone numbers, emails, etc. Therefore, the survey is completely anonymous.
3. Data Use.
The data is used purely for scientific research purposes and will be used for completion of a master thesis at the University of Amsterdam. No additional uses for the data will be made.
Thank you for your participation!
Section 1: Background
Please provide some information about your work background. 1. Please indicate your position within a company
Pažymėkite tik vieną ovalą. Ship Building Engineer Mechanical Engineer
Automation/Electrical Engineer Interior Engineer
Production Engineer Kita:
2. For how long have you worked in your current position? Pažymėkite tik vieną ovalą.
< 1 year 1 to 3 years 3 - 5 years 5 - 10 years > 10 years
Appendices
2/5
3. Do your work activities involve working with/complying to maritime rules/standards issued by
classification societies (Lloyds, B.V, others)? Pažymėkite tik vieną ovalą.
Yes No
4. How often do you work with maritime rules? Pažymėkite tik vieną ovalą.
Very often - daily Often - 3/4 times p/week Sometimes - once p/week Rarely - once p/month
Very rarely - once p/2 months or less
Section 2: The Use of Rules
In this section, questions deal with the use of rules/standards published by classification societies. More specifically, the questions aim to capture your experiences when using the rules. There are six question in this section.
5. How important are the rules in the work that you do? Pažymėkite tik vieną ovalą.
1 2 3 4 5
Unimportant Very important
6. How easy it is to use/comply to rules during your work process? Pažymėkite tik vieną ovalą.
1 2 3 4 5
Very easy Very difficult
7. Do you consider working with rules to be a time-consuming process? Pažymėkite tik vieną ovalą.
Yes
Somewhat time consuming No
8. Based on your experiences, do you consider rule compliance process to be a predominantly
manual process? E.g. finding and checking regulations must be done manually. Pažymėkite tik vieną ovalą.
Yes, the rule compliance process is based on manual work Some parts of the compliance process require manual work No, the rule compliance is not based on manual work.
3/5
9. How satisfied are you with the data formats (Word, PDF, Excel, etc.) in which rules are
delivered to you?
Pažymėkite tik vieną ovalą.
1 2 3 4 5
Unsatisfied Fully satisfied
10. How often do you need to check different files (rules) to get all the necessary information you
are looking for?
Pažymėkite tik vieną ovalą. Very often
Often Sometimes Never
Section 3: Rule Structure
The following questions inquire about the rule structure and your satisfaction with their composition. There are four questions in this section.
11. How satisfied are you with the way rules are structured/organised? E.g. organisation of
chapters and paragraphs, location of specifications, connecting/referencing different parts of a standard/rule, etc.
Pažymėkite tik vieną ovalą.
1 2 3 4 5
Unsatisfied Fully satisfied
12. Do you perceive information in rules often scattered/dispersed in different places?(E.g.
different chapters, paragraphs) Pažymėkite tik vieną ovalą.
Yes, information is often scattered in around Sometimes information is scattered around No, information is never scattered around
13. Do you think that better structure/organisation of rules would make them easier to use? Pažymėkite tik vieną ovalą.
Yes No
4/5
14. Information models are used for representation of concepts and relationships, constraints,
and rules among them. Information modelling is already widely applied in standards/rules in construction and engineering and has improved information management. Could information models improve the rule usability in maritime sector?
Pažymėkite tik vieną ovalą.
Yes, using information models in rules would improve their usability No, using information models in rules would not improve their usability I don't know
Section 4: Specification Compliance
This section deals with specification compliance - finding and checking specifications of different parts/objects in a rules. The questions aim to capture your experience on the structure and information provided by these specifications. There are six questions in this section.
An example of a specification: exhaust gas pipes and silencers are to be either water cooled or efficiently insulated.
15. During your work process, how much time do you approximately spend on finding and
checking specifications, outlined in a standard/rule? Pažymėkite tik vieną ovalą.
0-20% of my time 20-40% of my time 40-60% of my time 60-80% of my time 80-100% of my time
16. How would you rate connectivity/linking between different information sources? For instance,
how well similar specifications in different rules are connected/referred to? Pažymėkite tik vieną ovalą.
1 2 3 4 5
Very poor Excellent
17. How well-written are specifications you see in maritime rules? (consider SMART
specifications: specific, measurable, acceptable, realistic, time-bound) Pažymėkite tik vieną ovalą.
1 2 3 4 5
Poorly-written Well-written
18. How easy it is to find/locate specifications in rules? Pažymėkite tik vieną ovalą.
1 2 3 4 5
5/5
19. Which of the following problems/aspects do you recognize from your experience working with
specifications (and rules as a whole)? Please select all applicable options. Pažymėkite viską, kas tinka.
Finding and checking specifications takes a significant amount of time Majority of specification compliance work has to be done manually
Information is scattered: different sections or different rules/standards have to be used to find information on one specification
It is difficult to find applicable specifications for different products/objects
Throughout the product life-cycle (a life-cycle of a vessel for instance) specifications are often updated and thus, complicate the compliance process.
I do not recognize any of the problems above Kita:
20. Could specification compliance process be made easier if linking would be done between
different entities? E.g. linking (related) objects and their properties - linking/connecting data. Pažymėkite tik vieną ovalą.
Yes, linking different entities in rules/standards would improve the specification compliance process
No, linking different entities in rules/standards would not improve the specification compliance process
I don't know
The End of the Survey
Thank you for taking your time completing the survey. If you are interested in seeing the final results of this research, please provide your contact details in the box below (optional).
