Unifying project knowledge: Navigating the data-driven and cloud-powered frontier in engineering and production tools
Posted on October 01, 2024
This paper was first presented at ICCAS 2024 (International Conference on Computer Applications in Shipbuilding) in Genoa.
1. The digital transformation challenge
The last two decades has seen an increasing emphasis, through terms like Digital Transformation and Industry 4.0, to do more with data using IT technology and embracing automation. In the marine industry, this has focused minds on targeting inefficient data handing processes or improving the way the humans are supported by tools. A negative characterisation of this progress would show concern towards job losses, yet humans are essential to all aspects of ship design, shipbuilding, and operation. No-one likes to be involved in processes that are inefficient or have limited perceivable value. This has a detrimental effect on personal motivation and process quality. Tools have been an essential part of building ships since we started shaping fallen logs into efficient means of transportation. Until recently, humans have always led the development and use of tools but in the modern era it can feel that IT systems have a dominating position over the human element. Key systems such as CAD, ERP, MES, PLM, familiarised as three letter acronyms, have become essential to large scale shipbuilding projects yet it is not always easy to work between these systems and it can be difficult for smaller enterprises to embrace these tools without significant disruptions to business continuity.
Despite positive promotion, especially from the marketing of IT tool vendors, industry surveys show that only 5% of Digital Transformation projects have delivered complete success [1], with only 14% of managers (1733) agreeing that sustainable improvements were achieved (McKinsey Global Survey 2018). Reasons for this highlight; shortages in expertise and digital talent (31.4%), targeting cost reductions rather than as an investment in process improvement using digital technology (31%) and “general cultural issues” (31%) [2]. “Failing to take employees or team members on the digitalization journey” is another reason given for lack of success [3]. It seems that to ‘enforce change’ from company leadership in the name of the Digital Transformation is a recipe for limited success – experience collected from process improvement studies conducted over the last century indicates that change is more sustainable when it come from within the organisation.
Dig deeper into the principles and opportunities [4] offered by the Digital Transformation and it is not long before you come across LEAN, which itself can be traced back to the work of W. Edwards Deming and Plan-do-check-act, and further back to the concept of the scientific method which emerged from the work of Francis Bacon (Novum, Organum, 1620). In manufacturing, The Toyota Production System, is one of the best know examples of LEAN philosophy. While tools are important, improvements resulting from their introduction are often short term. The programme highlights that two-thirds of improvements are due to the engagement of the humans in the process and are essential to long term success and sustainability. They embody the expertise and knowledge required to make successful processes within an organisation. The philosophy is clear, tools should support the humans.
For the last two decades, the software and IT industry has also taken a LEAN journey. The introduction of Agile methodologies, such as Scrum, Kanban and Extreme Programming, have been codified into the concept of DevOps [5], an equivalent of the Toyota Production System. While most developers would understand the impact of DevOps through the introduction of processes to automate tasks, the key starting point is assembling the ‘team’ together to discuss their challenges, identify opportunities for improvements, targeting and reviewing them. It is an iterative and continuous process, which many will know from manufacturing described as Kaizen, the Japanese term for continuous improvement. The goal of DevOps is to “Reducing hardship and drudgery in our daily work through continual learning in order to achieve the organisations’ goal” and is something we would all happily embrace. There are many similarities between software development and ship design. Both are processes which create solutions to many unique challenges as part of building a much larger complex system that must be maintained in good condition for a long period. Shipbuilding began embracing LEAN almost a century ago, optimising manufacturing processes with the introduction of block assembly and pre-outfitting, for example. Improvements to engineering processes are also possible, but it cannot be achieved through the adoption of tools alone. The human part of the process must lead change for it to be sustainable.
2. Aren’t we already data-driven?
The origins of Information Management in shipbuilding can be traced back to the 16th Century [6]. In 1572, Mathew Baker became Her Majesty’s Master Shipwright (to Queen Elizabeth I) by engaging the “customer” with the first recorded use of paper-based technical drawings created by precision drafting tools. Prior to this, shipbuilding as a craft was handed down, father to son, master to apprentice, and based on the Italian or Portuguese styles. Consequently, ship design transitioned from what can be considered an evolutionary process to a revolutionary process [7], with early drafting tools providing the catalyst to create ‘virtual prototypes’ that could be reviewed against the opinion of contemporary experts. Fredrik Chapman would not publish the first naval architecture tools to evaluate performance quantitatively for another 200 years [8] whereafter calculations would have to be performed by hand until the introduction of computers.
In the 1960-70s computing technology, computational geometry and early CAD systems matured enough to allow representation of structural parts, equipment and systems, although most engineers would not have direct access to these tools until the emergence of desktop computing technology in the late 1980s, early 1990s. While representation of hull surfaces and ship arrangements in 3D were important evolution for design and naval architecture, the emergence of NC control was potentially a more significant development for shipbuilding. These digital instructions had to be compiled by hand, a tedious process. Many of the early ship design systems such as Steerbear (subsequently Tribon, then AVEVA Marine), AUTOKON, NUPAS (later Cadmatic) originated from the need to generate this information more efficiently. Today, these tools have evolved the ability to design complex structural arrangements, graphically, with virtual steel assemblies organised into manufacturing stages and efficiently producible parts. Digital cutting information extracted from the model is digitally sent to CNC machines including compensation to account for welding distortions. Annotation is directly marked onto the plate to assist assembly, eliminating the need for positions to be manually applied. This reduces the amount of supporting documents and manufacturing information required. Subsequently, digital data may be extracted from the 3D model to drive robotic welding systems.
This is a picture of data-driven ship production that has been possible for more than two decades. In large scale ship production, this degree of automation from design to assembly is essential to control quality and remain competitive. However, integrated manufacturing systems impose a constraint on the configuration of the design such that it should be producible using the available tooling. In addition to the weight and size limits of assemblies, the organisation of plates and stiffeners, for example, need to be optimised to match the capability of automated welding systems. Use of these automated processes is an outcome of focusing on LEAN manufacture. However, it does mean that the yard’s production facility becomes optimised to certain types of ships. A yard experienced in creating structure for commercial “box shaped” vessels with aligned stiffeners optimised to simplify production, for example, could not easily use the same tooling process for a naval vessel with structure optimised for shock.
Large proportions of assembly in shipbuilding, however, remain reliant on human effort. Almost all outfitting involves the assembly of uniquely shaped parts in unique areas. Beyond panel assembly there is little effective opportunity to use robotic welding due to the uniqueness of each weld or junction. For now, humans are far more effective at adapting to these configurations than machine. Human fabricators require information to direct them on what to build, what parts, tools and consumables they will need, when to perform the task and how to confirm the completed job meets specifications. It is possible to automate the generation of some of this information from the CAE system, but much of it will need to be produced by the production design team, coordinated by the planning team. They use their experience to interpret the CAE design and their understanding of yard practices to author manufacturing instructions. This typically consists of individual drawings, documents and files, potentially with several revisions of each, as inconsistencies between the design, engineering and productions processes are identified and intent is matured.
Small to medium sized shipyards have less opportunity to take advantage of automation. They may produce a greater variety of vessels, ships with special operational characteristics or offer a bespoke service or product. Competitiveness is achieved by being agile within the market, which requires being actively agile within business operations and processes. The removal of automation as a competitive advantage places a greater reliance on the human teams to work together and find efficiencies in project delivery. LEAN can be applied to the engineering process, for the benefits of project execution and manufacturing – just as it has been applied using DevOps to the software industry. The tools and opportunities for automation may be different but DevOps starts by asking the question of those involved – what can we improve? Given the high degree of reliance on delivering mature, actionable information to humans in shipbuilding, a focus on whether yard practices treat this commodity efficiently is worth investigation and investment.
3. Information management in shipbuilding
The CAD/CAE process focuses deeply on the configuration of the product (the ship), how its component parts function together and how information may be extracted to improve manufacturing efficiency. In moving towards production, a growing number of tasks require practical expertise supported by access to product detail and configuration information. These team members don’t want to be burdened with the additional challenge of having to learn complex software to occasionally access critical details. These disciplines may focus on project and production planning, project resourcing, procurement, finance, materials management, or time/task management with each requiring domain specific tools to create and manage the data that support these activities. Here we find ERP and MES systems with capability to cover these tasks across a range of different industries. As expensive tools, smaller shipyards often cannot afford to invest in these systems and choose to use alternatives such as spreadsheets and lists to capture the data required to drive their business. While practical and agile, relying on file-based data sources often creates challenges when scaling to larger projects due to the limitations of concurrent access and version control, resulting in duplication and time lost figuring out which is the latest version.
With such a wide array of different software systems and vendors used in complex shipbuilding projects it is challenging to have an integrated picture of the project. It is impractical to provide all personnel with expertise to access every system to build their own picture of the project and for data integrity reasons often necessary to restrict access to prevent unintended changes of datasets. Integration between systems is desirable and provides an opportunity to automate data exchange processes which are error prone when conducted manually. Bespoke integrations are typically costly to implement, often exceeding the purchase of the connected systems themselves and translating information from one system’s detailed datasets to those of another, respecting the yard practice and process, can take time to develop.
Given the challenges of integrating different systems together and managing the procurement process across multiple vendors, many shipyards IT purchasing departments focus on a strategy of embracing systems produced by a single vendor on the assumption that these will already be integrated. This approach favours headline industrial software vendors such as AVEVA, Bentley, Hexagon and Siemens who have, through acquisitions, a diverse range of engineering software products including many which focus on ship design, engineering, production and operation. There is the perception that all products in the portfolio will be integrated but that is not always the case. The shipbuilding industry is small and not as cash rich as, for example the Oil and Gas sector, and software specific developments tailored for this industry often experience reduced priority and resistance to innovation due to low return on investment. Smaller, domain-focused, marine software vendors retain enthusiasm and expertise, and consequently more agility to solve emerging challenges like integration far faster than the large vendors. All vendors must address integration but there is more than one way to achieve it.
3.1 Product Lifecycle Management (PLM)
Product Lifecycle Management (PLM) solutions have become the tool of choice for managing complex project information from design through to operations. PLM systems arose out of the early CAD systems that relied on files to store configurations of individual parts that make up larger assembles. Consequently, the need to control and understand the implications of change to this fragmented data was important to avoid version conflict and to allow those working on the project to identify and access the right content including information generated from the model, such as drawings. Today, CAD/CAE systems are far more integrated and rely on databases to keep information consistent, yet their remains a considerable amount of fragmented information associated with engineering projects.
PLM systems address the integration challenge by capturing datasets, files and documents, controlling versions, access, and how data is changed. It is attractive to interface software systems to PLM eliminating the need for manual data transfer. As these systems have matured, elements of ERP and MES capability have been included, supported by capability to communicate with production and operations processes beyond the office achieving real-time feedback, a goal of Industry 4.0. Growing with content the system becomes, by definition, the “Source of Truth” and other systems become subservient to its operation. By policing all data and how it is changed, these systems deliver order, control and consistency in project execution resulting in improved product quality, defect traceability, and greater repeatability. Consequently, PLM systems today offer far more than just Product Lifecycle Management and should perhaps be described more as Enterprise Information Management systems.
3.2 The applicability of PLM in shipbuilding
PLM has traditionally tasked with tracking the changes that happen to the engineering configuration as product performance is reviewed and component failures are discovered in operation. It sees signification application where large number of discrete or batch-type products are manufactured, such as food, chemical products, cars, aircraft and washing machines etc. Shipbuilding has very different product and production characteristics. Rather than build high numbers of repeated products with hundreds of components, where the configuration of each must be tracked in time to minimise the impact of defects, ships are individual products consisting of millions of components where the process of putting these together must be precisely orchestrated, but a step may happen only once. In the design phases there is far more focus on determining system configuration, and subsequently the component and integration configuration to meet owner requirements, supported by the information needed to build the vessel (manufacturing instructions). Few components in the product are precisely the same, limiting the opportunity to optimise manufacturing with production-line style automation, however, agility is achieved by focusing on the effectiveness of manufacturing steps that can produce components with a variety of geometric configurations, and efficiently assembly them together to ultimately form the ship.
Ship building also has some unique requirements rarely supported by PLM systems. Bill of Materials (BOM) are typically captured, which list the parts in a manufacturing task. However, in shipbuilding it is more typical to represent this in a tree structure which organises and sequences the joining of parts into assemblies, which are further joined into larger assemblies – a process often termed Assembly Planning. This provides additional understanding of, for example, weight and size, so that it can be managed against crane capacities, workstation and building entry opening size and another other material handling constraints that a yard may have. Junctions between parts and assemblies can be used to identify welds, and subsequently indicate the amount of work content and hours.
The ability to reuse successful engineering from prior projects is a highly desirable requirement that has yet to be solved satisfactory by any software system. Often termed Sisterships or Hull Applicability, the process supports the development of a principal class project, from which derivative vessels are built with minor alterations to account for learning during production, early operation of first of class vessels and equipment obsolescence. Furthermore, the opportunity to bring in successful equipment, systems, modules or configurations to a new design from any prior project, without losing supporting information would create both a productive and agile experience during early project phases. It would allow the design team to focus on addressing risks rather than spending time on regenerating definition to support evaluation. Since PLM identifies as the custodian of design information, these systems are ideal candidates to lead in delivering this capability.
In shipbuilding information management, due to lack of capability, PLM systems usually have an equal relationship with CAD/CAE, ERP and MES systems rather than being the lead system. Consequently, without a clearly defined Single Source of Truth to communicate project information and maturity transparently, it continues to be challenging to all teams and disciplines working in the project. This suggests that how the information is communicated to and accessed by team members could be a higher priority than focusing on clerical procedures used to store and control how project data is changed.
4.The engineer’s perspective
Information drives all engineers, especially in ship building. They need to understand the requirements and capability of systems to understand how to combine them in a confined space and understand how to configure them so that they can be manufactured. The ability to find project information and understand how it relates to the current question or task is imperative. While many information management systems promote themselves on the ability to deliver information to support decision making, the capability engineers themselves say they require to achieve a seamless understanding of the project is rarely documented. Why not ask the Shipbuilder? As part of their PLM implementation, Ulstein conducted an internal survey to understand what team members desire most of their information systems and was presented at the Cadmatic Wave Meeting in 2023, Figure 1.
Figure 1. Headline comments from an internal survey conducted by Ulstein presented at the Cadmatic Wave User Meeting 2023.
The feedback shows that time searching for information continues to be a dominant response, with other comments providing specificity:
- It’s not just about finding information. Recognition is given to the fact that information is stored in multiple systems and tasks such as detailed planning are completed by different teams with little visible between or across the project. A common way of sharing information is desired, such as a Single Source of Truth.
- It is difficult to judge the quality of data and ensuring quality takes effort.
- Understanding the maturity and status of engineering information is challenging and there is no process available for moving completed tasks forward in a workflow.
- It is difficult to share information in a way that can improve project quality, subsequently, knowledge remains with the individual.
Reframing the comments, it can be said that:
- Finding information is a first step.
- Information needs validation – help the user achieve this.
- Understanding the maturity of a piece of information and where it may be positioned in a workflow is important. The opportunity to review and approve information is highly desirable.
- The opportunity to add feedback to capture observations, objections, and experience is imperative.
5. Better access to project information
Finding information, i.e. searching, is a performance metric frequently used by software vendors in their sales delivery often leveraging survey data produced by International Data Corporation (IDC). Between 2001 and 2011, IDC conducted general surveys to understand the time people spent looking for information [9], finding that between 15-30% of the working day was used for this task. While an obvious opportunity for business improvement, there is little evidence across the decade surveyed that time searching for information was reduced. Since 2011, IDC focused on role and domain specific investigations. In engineering, we do not need a better Google, but better organising the way information is retrieved can accelerate access to relevant material.
Digital Transformation and Industry 4.0 reaffirms the need for feedback from processes to confirm they work as expected, going further by demonstrating how digital technology can be used to capture, store, aggregate and present quantified performance in real-time sourced from important business or production systems. In this respect, the need to expel effort to find data is eliminated by automating data capture and tailoring analytical routines to deliver intuitive presentations. These summarise pertinent information and support increasingly detailed breakdowns accessed by interacting with presentation graphics using the mouse or touch screen. In this respect, knowledge is gained by following a pathway along contexts of interest to the user. With an intuitive user interface, users can learn how to explore information available to them without need for training. Consequently, anyone with access to the presentation can get a picture of the situation regardless of whether they have the skills, expertise or training to access the domain specific software systems where the information is stored. A clear picture of a project is available regardless of whether you are a structural, piping, mechanical or electrical engineer, fitter or welder, or whether you are in a junior, senior, management or director role.
5.1 Creating a data-driven experience from fragmented documents and information sources
Bole [10] explored the opportunity for improving access to engineering and manufacturing information by utilising contextual relationships between the fragmented datasets and documents that are typically produced in a project. This information may be stored across the organisation on servers, in document management systems, databases and specific engineering software. By analysing attributes, data fields and text found in 3D Models, spreadsheets, databases, APIs, drawings, diagrams and documents for patterns that define key identifiers, values or tags used in the project it is possible to connect information together. Associations between specific types of document and data allow relationships between information to be classified. These classifications can then be used to structure the presentation of information on-screen enabling users to find knowledge important to them in an intuitive way.
5.2 Interfacing between information systems
The ability to automatically extract information from the principal project information systems is vitally important to create an integrated view of the project and exploit its data. Just as it would be ineffective to train the whole project team to use every expert information system to access the information they need, there is equally limited value to extracting entire datasets for display since this will slow down both software and comprehension of the information. Only summary information is required, enough to understand the situation. This data is typically easy to access and should not require deep understanding of proprietary data structures used in each system.
Today, it is common for information systems to have APIs which allow external systems access and to sometimes manipulate data stored inside proprietary software. SQL interfaces and Web/REST APIs provide standardised frameworks which define methods of accessing information avoiding the need to develop expensive bespoke interface between systems. It makes it unnecessary to replicate data into centralised control systems like PLM for comparison since it can be accessed directly and support automation of cross-checks of engineering attributes between systems. This is a good example of where DevOps thinking can introduce automation that both reduces inconsistencies and eliminates the tedious manual review work.
5.3 Intuitive configuration: a key enabler
A key enabler in the recent growth of information management tools which retrieve information from systems and aggregate it together is the ease in which they can be configured. Unless the tool is easy to learn, most are unlikely to sustain long term use given that they are usually selected based on added value rather than functional need. It is essential that the competency to configure these tools be maintained within the organisation to be resilient to staff-turnover or the need to rely on software vendor services. The growth of low-code or no-code techniques combined with API standardisation practically eliminates the need to write bespoke software interfaces. Although the need for good documentation to understand how an API is structured remains essential. Well-designed software configuration can accelerate adoption and given the ubiquitous nature of web-based applications, a professionally designed configuration page means that options can be easily understood and physical access to computer servers becomes unnecessary.
5.4 Defining the single source of truth
By bringing engineering data together in this way satisfies several solutions. It creates the ‘Single Source of Truth’, a system used across the organisation or project where data can be trusted, checked and referenced with colleagues. Where this aggregation does not physically store the information, the alternative term ‘Single Point of Access’ may be considered. Equally, in may be considered a Digital Twin capturing the current state of the project. 3D Models may be used to access all information associated with selected parts in the project, but these should not be considered essential. 2D drawings can serve as an alternative method of access where information is overlaid much like a GIS system. This approach can also support users who are more comfortable using 2D plans and are less familiar with the technology of navigating 3D models. In ship operations, where a 3D model may not be available, optical surveys like point clouds and virtual tours can be combined with overlays in 3D to provide an equivalent experience.
6. CADMATIC eShare – a solution for shipbuilding projects
eShare is a web-based application which aggregates 3D engineering models, documents, drawings and diagrams, point clouds and meshes, and datasets from external systems. Information can be filtered by hierarchical breakdowns and by shading the 3D model or document links through the categorisation of any data field within the 3D engineering model or external dataset. It is primarily a read-only solution but supports custom attributes which capture progress through a workflow, supporting maturity management, or grouping, to capture manufacturing packages for example. Furthermore, markup of views of the 3D Model and of 2D documents and drawings can capture observations and objections. Markups can be customised with workflows providing the opportunity to instigated change or feed into formal review process. Revisions to marked-up content can be compared with their original state at the time of capture.
The system has three principal objectives:
Consistency: Lack of alignment between data in information systems, in people’s knowledge and between organisations can lead to failures during manufacture and operation of the vessel, incurring cost and delay while rework or rectification is completed. Presenting a common view to as wide audience as practicable increases the chance of an inconsistency being highlighted especially if it can only be identified by someone with pertinent expertise. Consider including the customer (ship owner or operator), sales team and sub-contractors as part of the review team. Exposing inconsistencies as early as possible minimises the chance of incurring unforeseen costs.
Communication: Providing clear and precise project information on a continuous basis allows all project team members to maintain a common understanding how the project is evolving. Package the information so that when interacting with others they can direct attention to the specific point relevant to the communication. Provide capability so that points and decisions can be captured and subsequently, targeted to and worked on by the responsible experts.
Consolidation: Create an access point where, regardless of role, department or organisation, a comprehensible view of published project information and maturity can be found, controlled by appropriate access rules. Present data in-context with the users’ understanding of the project and provide them with intuitive tools to enhance detail on a topic when desired. Access information directly from source, either live or as published, to ensure continual trust in data such that confidence to take any decision or action based on the information may be maintained.
As a web-based application, the user-interface is subdivided into pages each with a key role:
- The Model View provides a view of the 3D model, a virtual representation of all physical parts in the product. On selecting the parts, relevant information gathered from all connected systems is presented such as data fields, document references, maturity status and associated comments or markup. Properties of the model can be reported using breakdown structures and dynamic shading based on the categorisation of any data field. The model may be “marked up” by capturing an image of the view, applying annotation and setting custom fields, and assigning it to a user or group to process.
- The Document View uses a hierarchy to break down the organisation of documents accessible from the system. Documents can be retrieved from the CAD/CAE system, from file systems or from a Document Management System provided by PLM Systems, M-Files or Microsoft SharePoint. The user need not care where a document come from and only has access to those their role allows. Different hierarchies may be used to organise access to documents, such as by system or by approval state. Documents may also receive markup, a view of the original document captured as an image, with annotation and custom fields.
- The Map View provides a digital General Arrangement of the ship. This allows navigation of the project from a deck perspective and support users who find navigating the project in 3D challenging. Smart points and 3D markup are displayed, with details presented when selected. The map view can also be used to direct the 3D viewing position to a particular point and direction.
Two further views provide the capability to search the model, with the option to display results in a grid format supporting engineering list views, and a dashboard style home screen which supports notifications of recent markup changes, saved engineering lists and histories of operational data.
6.1 Principal capabilities
Unlike CAD software where functionality is driven by specific engineering modelling needs, Information Management solutions are often configured around capability. A customer may choose to configure as much capability as they understand and need, extending it as they learn how to adapt the software to existing business processes, or optimise business processes based on the capability of the software. In eShare, the principal capabilities are described as follows:
6.1 (a) Accessing information in context
Bole [10] highlighted the simple concept of hyperlinking as a way of providing progressive access to discrete information sources. While a mature method for linking HTML documents that form the world wide web, they can be used to connect and classify associated engineering information. In eShare, 3D Models can link to documents and relevant text inside these documents. Drawings and diagrams can link to 3D Models or Smart Points, Figure 2. Documents published to eShare by the Cadmatic CAD/CAE system are automatically linked to 3D model part references while all other documents can be automatically linked by configuring rules (Regular Expressions) which match specific patterns of text to certain model attribute values, producing the link between the two pieces of information. Furthermore, model hyperlinks can be defined to jump to specific pages in other web-based systems allowing eShare to be used as an access point to more detailed information available in PLM, Planning or ERP systems. All connections are configured through the web-based administration pages and access to the server to perform typical administration tasks is unnecessary.
Figure 2. Hyperlinks used to connect data sources providing contextual discovery of relevant information
6.1 (b) Hierarchical breakdowns
Engineering projects are often subdivided into comprehensible groups of elements using hierarchies or breakdown structures. In many cases, the principal hierarchy is often defined by the way the project is stored in databases or files rather than by engineering needs. This can often constrain change to the project preventing consideration of alternative build strategies. In eShare, hierarchies are generated from model attributes, categorisations, and associations, Figure 3. Data attributes are typically values associated directly with the 3D Model whereas the categorisation offer a degree of customisation, using various querying syntax to convert string, date and numerical data into groups. Categorisation can be performed on data received from connected systems as well as 3D Model data fields. Associations are special relationships between model items and structures such as spatial arrangements or welding information exported from the Cadmatic CAD/CAE System.
Each level in a hierarchy may be defined from multiple attribute values, categorisations, or associations. A hierarchy could start with a block breakdown, then schedule dates from a planning system, then delivery status from an ERP system. Using this approach a user may be able to visualise critical project status information projected onto the 3D Model, effectively a Digital Twin of Production Status.
Figure 3. Breakdown structures can be generated from combinations of model attributes, maturity information and data extracted from external applications. Hierarchies can be rapidly defined from the configuration page.
6.1 (c) Visual styles
Colour shading of the 3D model provides an alternative method of presenting a project breakdown, albeit with only one level of categorisation, Figure 4. Visual styles are be defined from same categorisations used to define levels of hierarchies. Hierarchies and Visual Styles may be used together to achieve cross filtering, where two metrics are used together to understand information.
Visual Styles may also be used to colour the hyperlinks to model items in documents and drawings. This allows users who may primarily access information using documents such as the Production team to see a visual indication of the state of model information such as production readiness.
Figure 4. Visual Styles dynamically shade both the 3D Model and document Hyperlinks conveying project information, plate thickness (left), nesting status i.e. nested or cut (centre), and engineering maturity status shading drawing hyperlinks (right).
6.1 (d) Capturing maturity workflows and work packaging
Many of information systems used in engineering often provide the capability to track the maturity of components through a workflow. If a dedicated system, like PLM, records maturity and is accessible for all team members, values can be accessed by eShare using an API and presented as part of a hierarchy or visual style. However, if such a record doesn’t exist or is not accessible to all team members eShare can provides this capability using Status Trackers.
Operating in Workflow mode, a Status tracker represents a sequence of states between which valid transitions are defined, for example, through the maturity stages of engineering, review, and production, with back step transitions provided to address rework. Certain states and transitions may only be available to specific disciplines or lead users, controlling who can approve change ensuring, for example, work does not enter production before it has been reviewed.
Status Trackers also support open operation where there is no workflow. This supports the capture of engineering components into groups supporting basic work-packaging or assembly planning, but without any hierarchy.
6.1 (e) Capturing observations and initiating change
It follows on from visualising and capturing maturity, it should be possible to register comments when something isn’t right and needs changing. While it may be typical to capture these during formal review processes, since the system is always live, observations now can be captured at any time, by anyone with access, creating the opportunity for continuous feedback. This agile approach to engineering review may be adopted as a culture, with appraisals happening continuously between the engineer and team lead, for example. Formal review meetings can be shorter, with focus on the important conflicts that need a wider range of experts to resolve but not disrupt their day more than necessary. Furthermore, by supporting mobility, the system can be taken to the work site and comments captured alongside photographs and video, bringing comprehensible feedback back into the engineering project.
eShare provides capability to annotate and provided custom fields on the markup of the 3D model and 2D documents and drawings. An image of the target captured at the time of markup creation allows it to be compared with later revisions of the content, supporting confirmation that the change has been made. By associating workflow, priority and responsible team or individual, it can be tracked through to completion. As annotation and redlining of drawings are database entries, any markups created during formal document and drawing review become data driven. Markups in 3D automatically reference part information supporting the generation hierarchies, Figure 5, and visual styles through which inconsistencies in the engineering can be managed and allow the project to grow in maturity in a controlled manner. Equally, markups can be used to instigate requests in the formal change management tracking solutions that may be found in PLM Systems or other project management software.
Figure 5. Markups and Status trackers information can feed into breakdown structures extending the data-driven experience to drive the project review process.
6.2 Application to different shipbuilding activities
The building blocks of capability described in 6.1 are applicable throughout the entire shipbuilding process. Although the information content at each stage may change composition from 3D in design, 2D in production and imagery in commissioning. eShare handles these consistently to ensure that all involved in the process can continue engaging between each other in the delivery of the project. Focusing on the specific applications of eShare in each shipbuilding stage:
6.2 (a) Engineering
During the engineering phase of a project, team members are primarily concerned with checking the design meets functional requirements, with the opportunity to register observations and objections. Understanding the maturity of information is important since, if configuration is in progress then it is premature to register comments, and if approved for production then engineering change becomes difficult and must be more critically justified.
Manual checks rely on the visual review of numerical and graphical information. Effectiveness is influenced by experience, stress and motivation. It is not unusual to have to check information authored across different systems, introducing tedium which results in fatigue. Automatically bringing information together to check in the same environment minimises this and allows the reviewer to identify subtle inconsistencies that might not be picked up if configurations are viewed separately. Opportunities exist to automate checks and are best applied to validate the most tedious of data consistency issues. Where multiple information systems are used in the engineering environment, having the ability to capture maturity, in a workflow and objections in a common and easily accessible system reduces the need to train team members in specific systems used across the project.
Typical uses within the Engineering Process include:
- Checking the format of key identifying attributes from 3D Model and external systems to confirm that the pattern of the text is as expected. Since identifiers are often manually generated this capability highlights any that have been incorrectly typed or have been initialised with a generic value.
- Confirming synchronisation data value between independent systems. With different software tools being favoured by specific disciplines but a frequent need to represent the same engineering attributes there is a significant chance of mismatch if the only way to validate this information is through manual review. By digitally reading and comparing values between multiple systems, differences can be efficiently highlighted using visual reporting, Figure 6. A typical example is between Diagram/P&ID schematics and equivalent 3D Pipe Model data.
- With the growing use of Multi-CAD environments and Point Cloud Surveys, comparisons can be made to confirm that 3D engineering models have the correct alignment. This can be approached both through a manual visual comparison or using a clash checking process to automate and track inconsistencies.
- As manufacturing starts but a large proportion of engineering design is still to be completed, visual reporting can feedback when engineering components begin their manufacturing journey and start to be installed. If changes need to be made, understanding the current situation reported by production teams can highlight what changes are acceptable by accurately accounting for any cost and rework delays.
- Within the large number of engineering components in a project sometimes the smallest elements are most critical, such as penetrations. The breakdown structure can be used to filter out unimportant elements to focus on these physically small elements to understand where they are located around the vessel. Status Trackers can capture management characteristics such as their Watertightness, intended production process and production status. This is essential for coordinating between the different disciplines that interact with these components.
Figure 6. Comparisons and inconsistency highlighting between, (left), numerical attributes from 3D Model
and an Excel data source, and (right) of pipeline names in the Diagram and 3D Model.
6.2 (b) Production
Once engineering has signed off a configuration, production will organise and track the manufacturing process. Domain specific tools that capture tasks, timescales and supporting manufacturing information act upon parts and assemblies designed in 3D but rarely do these systems provide any visual context to understand what is going on beyond the identifying name. Overlaying this data on to the 3D Model provides a visual indication of the construction sequence, often called 4D Planning, and may highlight fabrication challenges such as lockout, where beyond a certain point it may be impossible to fit a piece of equipment without dismantling already completed work. Furthermore, by combining planning with delivery information it is possible to confirm whether imminent production tasks can proceed, or if not, alternative strategies can be investigated.
Another challenge experienced in production is the emergence of information that may not have a direct relationship to an engineering part, yet still needs to be easily found. These often take the form of discrete documents such as common work instructions, sketches, notes and delivery information, and subcontractor information. It may not be possible to associate this directly to engineering items, but associations can be made to blocks, compartments, assemblies, or custom spatial references.
Typical examples include:
- Using Status Tracking as a grouping capability to capture parts associated with a fabrication work package. Since most planning systems focus on scheduling and sequences the ability to understand what parts look like and any limitations that may come from installation in confined spaces can prevent unexpected situations before they happen.
- Being web-based, eShare can be made accessible to sub-contractors enabling them to feed in their own manufacturing and dispatch status, this workflow can be extended down the logistics chain. Customers have experimented using RFID tagging to capture parts at each stage without the need for manual entry. Subsequently, this status can be captured as attribute values and reported using hierarchies or visual styles.
- Confirming that manufacturing information for work packages is complete by generating breakdown structures that group documents to the work package name. Reviewing the documentation available against each work package confirms readiness to proceed.
- Through visual styles the current maturity status is graphically reported. Items urgent or pending can be easily identified and tasked for review and approval ahead of pending manufacturing deadlines. Routine cases can be easily progressed without the need to wait for larger team meetings resulting in a far more flowing and continuous delivery process. The breakdown hierarchy can be used to further filter on deadlines and delivery status to consider any pertinent factors that may drive decision making.
6.2 (c) Post-production
After manufacturing has been completed maturity again becomes important. Confirmation of commissioning and testing can be tracked, Figure 7, with documented outcomes, although final sign-off could be the more important transaction in the case of billing. At this stage, sub-contractors and external organisation may benefit from access to the project information system. Deploying eShare beyond the boundaries of the organisation, with appropriate security, provides them with access without involving or requiring any IT infrastructure competency on their part. Work completions may still involve paper signoff which in a modern project needs to be captured digitally.
Mobility, the ability to take the digital twin onboard, becomes desirable allowing ‘live’ updates of the project. It is essential that Mobile solutions have some capability to work without a live connection since internet access is not always available and usually non-existent inside the metal compartments of a ship. Taking the Digital Twin to the workface allows tasks to be supported by manufacturing instructions and technical documentation. Maturity and observations can be captured immediately with the opportunity to include media such as photos and videos which is easily captured by mobile devices. Status trackers may be used to capture installation, commissioning and signoff process which, as strictly non-manufacturing activities, may not be captured by an ERP system.
Typical examples include:
- Commenting can capture inconsistencies and observations, supported by photographs. With an associated workflow, inconsistencies can be managed and proactively closed.
- Documents may be produced that do not have a direct relationship to engineering components and there may be no contextual value representing the associations. Smart points allow sketches, signoff documents and notes to be related to a spatial position as an alternative context to the 3D Model, Figure 7.
- Use of status trackers to capture approval of penetrations during Class surveys has been prototyped, supported by the 3D model and drawings. The experience highlighted the additional challenges that occurs when there are differences between as-designed and as-built.
Figure 7. Linking physical paperwork generated during manufacturing into the 3D environment using spatial references, i.e. SmartPoints, when no existing model item is available to target the connection.
6.2 (d) Business reporting
The capture of maturity changes and observations creates a historic record of progress. This information can be useful to the business and engineering leadership to not only understand how overall project progress is performing but also support business performance comparisons by breaking progress down by team, system, compartment or block. Indeed, reviewing this information may expose a challenging engineering situation in the project either due to a reduction in the pace of maturity changes and increases in markup, comments or objections. Tracking this information may highlight a situation before team members are even aware of an emerging situation, especially if the team involved uses distributed working. This monitoring could be used to instigate team meetings to ensure alignment on the challenge before it impacts on schedule.
Dashboarding has become the accepted means of reporting business performance and intelligence, Figure 8. There are numerous software tools delivering this capability and eShare can be used as a data source. Many dashboards are web-based and creating the opportunity to hyperlink from dashboard content into eShare content to understand detail in context with other project information.
Figure 8. Microsoft PowerBI dashboards driven by data published by eShare showing, left, approval status for penetrations filtered by compartment, and, right, time history of markups opened and changed.
6.3 Implementation and adoption
For many information management systems, implementation and adoption is a long-term process especially when tailoring a PLM solution. That is not the case with eShare. Implementation can be rapid due to the ease in which it can be configured. Adoption within the organisation can grow with rate of understanding and knowledge the team members can sustain, adapting to their workload and needs.
New projects can be configured and published within minutes. Using templates and configuration, rules from previous projects can be applied directly to a new instance. Greater success is achieved if the organisation already has a mature and consistent naming standard for engineering components, drawings and documents. Configuration can be copied from one project to another as human readable text definition, supporting adaption if there are minor differences in project setup. This minimises the need to invest time manually preparing a new project before it can be used. Consequently, there is no penalty to create projects to prototype ideas, investigate new connectivity to external systems or trial new workflows. If successful, the relevant configuration can be copied to active projects, and if not, the project can be deleted. Customers have been able to implement eShare and deploy their project to both ship owners and refit yards to bring them into the engineering review process within a couple of days [11].
Just the publication of the 3D model file alone provides more capability that the equivalent standalone software eBrowser, as all approved team members can access and begin reviewing the project using their web-browser without the need to deploy project files or executables. Subsequently, the platform can now be used support DevOps improvements by facilitating internal discussions about what information they would like to see available and form a plan for implementation, with guidance, support and implementation by the Cadmatic Team if required.
7. eShare at Lurssen: a case study
Lürssen-Kröger Yard, based in Schacht-Audorf in Germany on the Kiel Canal, part of the Lürssen group, is a mega-yacht builder of vessels in the range of 50-120 m. In 2021, following long term use of Cadmatic’s 3D Model viewer software eBrowser, Lürssen implemented eShare allowing it to publish project content across the organisation without the need to distribute software or model files. In Spring 2024, the reach of eShare was extended by adding a small number of eGo installations on laptops, receiving positive feedback after 6 weeks of use from users typically involved on the production floor. Furthermore, several studies have been conducted to investigate the possibility of using AR/VR to support production processes based on eShare project content.
The initial objective was to capture the installation state of engineering components, bringing this information back into the planning environment, providing greater accuracy in both the management of production priorities and reporting the state of task completion to project leadership. Since much of installation involves the management of subcontractors, the ability to understand the state of work, quality of delivery and sign-off is important to business operation continuity.
The current ERP system comes from an automotive background. Many team members find it challenging to get to the information they need. Some have developed a negative view of the application. By connecting the ERP system to eShare, employees in production can now view the most important component information, such as the installation timeline or delivery status, all in one place. Another major step is the visualization of the parts lists from the assembly work packages in the ERP software in the 3D model. The components to be installed, which were previously defined in meetings, are visualized here. The planning and delivery information from the ERP software, which is displayed in the hierarchy in the 3D model, makes it very easy to check if the work can be done or not. The status tracker can then be used to obtain a very high level of information about the assembly status. This integration provides the planning team with a higher level of confidence in the accuracy of the information generated by this process.
Use of eShare now extends to the engineering teams. The decision to make changes is now informed by the installation state of engineering components offering agility to respond to late design requests. If change is necessary and impactful, production can be pre-warned of rework and resources adjusted based on the new information. The system can be used to engage with the production teams using the 3D model as a reference, reducing the degree of surprise when change requires removal of in-progress or completed work.
Dashboarding is used to report the installation status to the project leadership, powered by information inside the eShare project. Originally, this was based on information collected by the ERP system covering the status of materials and manufacturing tasks but excluded testing, commissioning and sign off. This significantly underestimated the completion state of the project. A more accurate review is produced with the combination of information available in eShare. Data is reported from eShare using the search capability to generate an Excel spreadsheet and can be filtered by Subcontractor, System, Areas and Compartments to produce a detailed understanding of progress and deployment of resources across the project. Presently, this is a manual process and Excel is used for Dashboarding since most engineers have the skills to understand how this works. It is more than feasible to automate this process using for example, Microsoft PowerBI, which although easy to use and learn, can require conscious effort to maintain a working knowledge of the configuration.
Deploying engineering information to the production workface can often bring surprising outcomes when ‘as designed’ is compared with ‘as built’. This provides some of the final opportunities to look for inconsistencies, especially when being independently reviewed by Class. Review of penetrations, for example, traditionally requires the use of many large format drawings which are difficult to review, especially in confined spaces. Lürssen is pioneering the use of augmented reality to deliver production reviews without the need for drawings. Microsoft Hololens technology is used to project the 3D model onto the partially completed ship environment. It takes around 10 to 15 minutes preparation to align the 3D model to the vessel, with the review of each deck typically taking 1 to 1.5 hours. In the future, the status will be updated from inside the AR environment and project updated immediately when Wi-Fi is available. Use of AR in partially built vessels is not without risk and it been found that the AR headset user requires support from an additional person to avoid holes in decks, trip hazards and other obstacles that could cause injury.
eShare is in operation on a continuous daily basis, supporting planning in connection with the ERP systems. It has increased the ability to find information and team leads have a much greater understanding of project state, especially those working in production. Since eShare has predominant enhanced communication and clarity, it is difficult to put a value on costs saved. 80% of users are happy with the solution, although the need to capture production status during the introduction of the system was perceived as an overhead. A progressive approach to applying the capability of eShare to projects, choosing to tackle one problem at a time very much in keeping with the philosophy of DevOps.
7. Summary
The impact of the Digital Transformation and Industry 4.0 is that organisations think more about how to exploit data to improve their business. Shipbuilding is human resource intensive where efficient communication of project information is essential to success. Individual documents continue to be important, but the propagation of data-driven integrated engineering, procurement and production tools supports a seamless flow of information around the business. These tools are often designed for the specialists that use them and limit access to those without familiarity, training or access rights. This restricts communication and could result in project inconsistencies and unforeseen costs.
As data models become more complex, the PLM philosophy of extraction, centralisation and control of data and becomes less effective as information is often only meaningful in originating software. With many information systems now providing APIs based on standardized frameworks it’s no longer necessary to access software through the UI or create bespoke interfaces, data can be read immediately. Cadmatic eShare uses these interfaces to read data and documents creating a unified picture of a project accessible via web-browser to teams and departments across the project. Capture of maturity and comments allow progress to be understood and objections highlighted. This aligns teams and minimises rework, directly impacting on unforeseen costs and delays.
A well designed, low/no code approach to setup configuration from the web-browser supports rapid implementation and minimises the need for the typical specialist learning required when adopting new IT software. Even a setup consisting of a single 3D model delivers value around the business. Incrementally, the team can begin to consider what project information and processes to integrate into the environment, and which teams will benefit. This people driven approach provides time for new capability to be adopted across the project and results in a non-disruptive and agile experience when delivering a Digital Transformation.
8. References
1. GRUBENMANN, M., ‘Ders Mench in der Digitalen Transformation’, https://innoscope.com/der-mensch-in-der-digitalen-transformation, 2019.
2. SOLIS, B., LITTLETON, A., ‘The 2017 State of Digital Transformation’, Altimeter, October 2017
3. EIGNER, M., ‘System Lifecycle Management’, Springer, 2021.
4. KERSTEN, M. ‘Project to Product: How to Survive and Thrive in the Age of Digital Disruption with the Flow Framework’, IT Revolution, Portland, Oregon, 2018.
5. KIM, G., HUMBLE, J., DEBOIS, P., WILLIS, J., FORSGREEN, N. ‘The DevOps Handbook: How to Create World-Class Agility, Reliability, & Security in Technology Organizations’, 2nd Edition, IT Revolution, 2022
6. JOHNSTON, S., ‘Making mathematical practice: gentlemen, practitioners and artisans in Elizabethan England’ Ph.D Thesis, Cambridge, 1994.
7. ALEXANDER, C. ‘Notes on the Systhesis of Form’, Harvard University Press, 1964.
8. CHAPMAN, F. H., ‘A Treatise on Ship-Building’, 1760.
9. WHITE, M., ‘Time spent searching – a chronology of the myth and some recent research’, https://www.linkedin.com/pulse/time-spent-searching-chronology-myth-some-recent-research-white/, 2020
10. BOLE, M., ROUSEAU, E., POWELL, G., ‘Taking Control of the Digital Twin’, SNAME Maritime Convention 2017, Houston, Texas, 23-28 October 2017.
11. ‘Glosten's innovative ferry retrofit project with CADMATIC eShare’ https://www.cadmatic.com/en/marine/references/glostens-innovative-ferry-retrofit-project-with-cadmatic-eshare/
ACRONYMS
API Application Programming Interface
AR/VR Augmented Reality/Virtual Reality
BOM Bill of Material
CAD Computer Aided Design
CAE Computer Aided Engineering
CNC Computer Numerical Control
DevOps Development Operations
ERP Enterprise Resource Procurement
GIS Geographic Information System
IT Information Technology
MES Manufacturing Execution System
NC Numerical Control
PLM Product Lifecycle Management
REST Representational State Transfer
SQL Structured Query Language
9. Authors' biographies
Marcus Bole is a business development consultant within the Information Management team at Cadmatic and helps customers improve productivity using eShare and associated productions. A Naval Architect, after several years delivering SOLAS ’90 and Stockholm Agreement stability upgrades to Northern European ferry fleet choose to specialise in software tools for shipbuilding with roles in development, support, training and business development. Specialisms include hull surface design, Naval Architecture design and analysis, ship and submarine manoeuvring, passenger ship evacuation, weight estimation and the hull steel design and production processes. For more than 20 years, he has developed and published the free naval architecture design and analysis software PolyCAD.
Niklas Nebel is a mechanical engineer who works as a production controller at the Lürssen-Kröger shipyard. He has made his way from programming CNC machines, to production planner and assembly manager in a defence company, to his current position at the Lürssen-Kröger shipyard. In addition to the day-to-day tasks of a production controller, he is the project manager for eShare at the shipyard and is also intensively promoting the topic of augmented reality at the shipyard.
M Bole, Cadmatic, UK
N Nebel, Lürssen-Kröger Werft GmbH & Co. KG, Germany