Laser scanning in manufacturing
industries
The potential and usability of laser scanning for industrial applications
JOHAN SVEDBERGER JONAS ANDERSSON
Master of Science Thesis IIP 2013:x KTH Industrial Engineering and Management
Abstract
Making mistakes or discovering errors too late in the factory layout process is very costly. Unfortunately, the layouts aren’t always accurate or updated which creates a degree of uncertainty when it comes to installation of new equipment and rebuilding facilities. It also leads to a lot of waste in movement when employees has to go out in production to perform measurements, take pictures and take notes in order to remember important details to avoid errors.
Lasers in land and engineering surveying instruments have been widely used for the last 30 years. A natural development has been to add a scanning mechanism to a total station that were already equipped with laser rangefinders and angular encoders, allowing automated measurement and location of thousands of nonspecific points.
The automobile industry has begun to see the potential of laser scanning, mainly because of the development of the software handling the scan results, the point clouds. Scania, in collaboration with the FFI research project at the Royal Institute of Technology (KTH), therefore wanted to investigate how the new possibilities of 3D laser scanning can facilitate the development and maintenance of production systems and how it could be implemented in the current factory design process.
By scanning three locations at Scania related to machining, assembly and aftermarket service the usability of the results has been investigated with the software Faro Scene and Bentley Pointools V8i.
The results of the study showed that the laser scanning technology can generate several possibilities for manufacturing industries. The technology can be profitable and the modern point cloud software and applications could support the work of the layout design process greatly. Three main application areas found:
Visualization & communication: The point cloud is an excellent information carrier and can easily be used as a visualization aid for meetings or simply to refreshing memories of a location. It also provides the possibility to view and examine a location remotely.
Gather information: The measuring possibilities are immense, allowing single point and distance measurements without the concern of interfering with objects. The method can to some extent replace the current approach in measuring buildings and floor flatness.
Simulation & verification: Software can perform advanced simulations and verifications of existing and future layouts, models and installations. Parts of the point cloud can be colorized, hidden, removed, duplicated or transformed. Existing 2D layout drawings or 3D models can be attached and verified relative the point cloud. The attached objects can be simulated with clash collision or differencing.
Sammanfattning
Att göra fel eller upptäcka fel sent i fabrikslayoutprocessen är mycket kostsamt. Layouter är inte alltid korrekta eller uppdaterade vilket kan skapa en viss osäkerhet vid installation av ny utrustning eller ombyggnationer. Det leder också till onödiga förflyttningar då anställda måste gå ut i produktionen för att utföra kontrollmätningar manuellt genom att ta bilder och föra anteckningar för att komma ihåg viktiga detaljer och undvika fel.
Lasrar inom mark- och lantmäteriinstrument har använts i stor utsträckning under de senaste 30 åren. En naturlig utveckling har varit att lägga till en skanningsmekanism till en total station som redan är utrustad med laseravståndsmätare och vinkelgivare vilket möjliggör automatisk mätning och lokalisering av tusentals ospecifika punkter.
Fordonsindustrin har börjat att se potentialen med laserskanning, främst på grund av utvecklingen inom hanteringen av punktmoln, resultatet från skanningen, på mjukvarusidan. Scania, i samarbete med forskningsprojektet FFI vid Kungliga Tekniska Högskolan, ville därför undersöka hur de nya möjligheterna inom 3D-laserskanning kan underlätta och stödja utveckling och underhåll av produktionssystem samt hur det skulle kunna användas i den nuvarande fabrikslayoutprocessen.
Genom att skanna tre platser på Scania tillhörande bearbetning, montering och service (eftermarknad) har användbarheten av resultat undersökts med mjukvarorna Faro Scene och Bentley Pointools V8i.
Resultaten av studien visade att laserskanning kan skapa många nya möjligheter för tillverkande industrier. Tekniken har visats sig vara lönsam och funktioner i moderna programvaror kan stödja arbetet för fabrikslayoutprojektering väsentligt.
Tre huvudsakliga användningsområden hittades:
Visualisering och kommunikation: Punktmolnet är en utmärkt informationsbärare och kan enkelt användas som en visualiseringsstöd för möten eller för att komma ihåg platser. Det ger också möjlighet att se och undersöka en plats på distans.
Samla information: Mätmöjligheterna är goda. Punkter och avstånd kan mätas utan att påverkas av skymmande objekt. Metoden kan även till viss del ersätta det nuvarande arbetssättet med uppmätning av byggnader och planhet av golv.
Simulering & verifiering: Programmen kan utföra avancerade simuleringar och verifieringar av befintliga och framtida layouter, modeller och installationer. Delar av punktmoln kan färgläggas, gömmas, tas bort, dubbleras eller flyttas. Befintliga 2D-ritningar eller 3D-modeller kan importeras och kontrolleras relativt punktmolnet samt att kollisionstester kan utföras.
Nomenclature
CAD Computer Aided Design CAM Computer Aided Manufacturing
FOV Field Of View
GUI Graphical User Interface
HD High Definition
IPS Industrial Path Solutions LiDAR Light Detection And Ranging PDM Product Data Management PLM Product Lifecycle Management
POD Point Database
PTL Project file
1
Content
1 Introduction ... 3
1.1 Background ... 3
1.1.1 Current problems at Scania ... 3
1.1.2 Origin of the thesis ... 4
1.2 Purpose ... 5
1.3 Problem analysis ... 5
1.4 Delimitations ... 5
2 Methodology ... 7
2.1 Method model of the study ... 7
2.1.1 Finding output to RQ1 ... 7 2.1.2 Finding output to RQ2 ... 7 2.1.3 Finding output to RQ3 ... 8 2.1.4 Finding output to RQ4 ... 8 2.2 Data collection ... 8 2.2.1 Document study ... 8 2.2.2 Interviews ... 8 2.2.3 Observation ... 8 3 Theoretical framework ...11
3.1 Digital factories & digital manufacturing ...11
3.2 Factory layout ...12
3.2.1 Layout software evaluation ...12
3.2.2 Factory Design Process ...14
3.3 3D laser scanning ...15
3.3.1 History and background ...15
3.3.2 Terrestrial laser scanner ...15
3.3.3 TLS technology ...16
3.3.4 Quality of TLS ...17
3.3.5 Bentley Pointools V8i ...21
3.3.6 Conversion of point cloud to CAD format ...25
4 Results of case studies ...27
4.1 Benchmarking ...27
4.1.1 Laser scanning at Volvo Cars ...27
4.1.2 TLS manufacturers and processing software ...29
2
4.2.1 Scanning locations description ...30
4.2.2 Scanning time study ...32
4.3 Usability of the software tested...33
4.3.1 Faro Scene ...33
4.3.2 Bentley Pointools V8i ...36
4.4 Cost calculations ...42
4.4.1 Expenditures ...42
4.4.2 Earnings ...43
4.4.3 MIKA calculation ...44
4.5 Factory Scanning Process ...44
4.5.1 Perform scanning ...45
4.5.2 Perform processing ...47
4.5.3 Use point cloud ...49
4.6 Factory Scanning Process implemented in the Factory Design Process ...50
4.6.1 Pre-study ...50
4.6.2 Project planning ...50
4.6.3 Final planning and installation ...50
5 Analysis and discussion ...53
5.1 Technology and applications of laser scanning ...53
5.1.1 Technology of laser scanning ...53
5.1.2 Usable features of the software ...54
5.1.3 Conversion to CAD models ...55
5.2 Cost analysis ...55
5.3 Scanning process model analysis ...56
5.3.1 Factory Scanning Process ...56
5.3.2 Factory Scanning Process in Factory Design Process ...56
6 Conclusion ...59
6.1 Result of the research questions ...59
6.2 Recommendations to Scania ...60
7 Future work ...63
3
1 Introduction
This chapter provides a background for the master thesis research and its purpose. It also involves the problem analysis, delimitations and the outline of the thesis.
1.1 Background
Today the global market puts higher requirements on exports and more production are sent to low-wage countries. The production and manufacturing industries requires ever-greater productivity, more savings and less waste. The need to design and build new factories or make changes to an existing factory layout has increased mainly because of the rapid changes in customer demand for both product quantity and product variety. In order to remain competitive and meet market requirements, businesses need to be more agile to plan, design and reconfigure the factory layout quickly (Chen, et al., 2012).
1.1.1 Current problems at Scania
Making mistakes or discovering errors too late in the factory layout process is very costly. The layouts aren’t always accurate or updated and the layouts accessible by the Scania production engineers often lack enough information of the ceiling in the buildings (Andersson, 2013).
Documents such as CAD or drawings often exist for most buildings, but these can be hard to obtain, difficult to understand without the proper context and might be changed without a proper update. This leads to a lot of waste in movement when employees have to go out in production to perform control measurements and remember important details to avoid errors. Some employees even has to photograph locations and take notes of measurements each time a new project begins, which also might be impossible when production isn’t down (Jardemyr, 2013).
The software used to draw and develop layouts at Scania is called LayCad , a tailor made version of AutoCad Architecture and it is mainly used by production engineers and facility engineers. In LayCad, the layouts are often represented as top viewed 2D incision just above floor level with some equipment and machines represented as 3D sketches. Moving parts such as doors and hatches also needs to be drawn opened (Allard & Sättermon, 2002).
The production engineer at Scania has the responsibility to make sure that the layout, covering each engineers responsible workshop area and common areas in the factory, is updated and representing the current situation, which should also be controlled once a year. The engineer is also responsible for maintaining the file structure, applying correct status of the layout file and purge obsolete files. Apart from each production engineer’s responsibility of the personal area, a layout coordinator has an overall responsibility of the whole production area consisting of the others’ layouts (Allard & Sättermon, 2002). DynaMate, a subsidiary to Scania, offers technical services within production maintenance & facility management. Their special competence includes electricity, ventilation, construction and robot installations (Scania Inline, 2013). All construction changes of the facilities at Scania must go through DynaMate to assure that the digital construction drawing is updated and making sure that all layout drawings have a common building origin to enable comparisons of the drawings (Karlsson, 2013). When the origin of the building is not marked out, entrepreneurs are hired to measure the facility. However, the majority of the 2D and 3D digital production layouts do not include information and documentation about the ceiling, ventilations and media installations, making it hard to gain an overview of the area without studying the actual location in detail, on site.
4
pillars are moved, new walls are built and other major changes are made without the digital layout being properly updated (Mårtensson, 2012). To subsequently update the CAD files manually would imply a major workload from control measuring the facilities.
Several of Scania's older buildings are only documented on paper drawings which have been scanned to pdf-format and therefore cannot be imported and compared in LayCad. The actual digitalization to LayCad of the scanned paper drawings are done stepwise when parts of the facilities are being rebuilt and are based on the old drawings and new measurements.
The uncertainty of how the reality differs from the layouts and drawings creates many problems and unnecessary work. A general perception of the production and facility engineers is that the layout basis lacks of information and is inaccurate, which obliges them to control measure and collect additional data manually using measuring tape or laser distance meters (Andersson, 2013; Johansson, 2013
)
. Unevenness is excluded in the layouts and in many cases cable shafts and other ditches are not represented at all. Creating new layouts, facility constructions or installing machines therefore demands knowledge of the actual site in order to avoid errors or problems since many details aren’t documented.Commonly, the production engineer has to go between the office and the production floor multiple times to measure and control the layout during a project (Andersson, 2013). The collection of measurements may interfere with the production which might need to be stopped. The results also need to be critically questioned and validated since the quality and accuracy of the measurement depend on the usage of the hand held laser distance meter.
With the overall responsibility of Scania’s facility service, DynaMate needs to travel long distances between their office and the locations needed to modify to collect information. The area is documented with several digital photographs and the interesting measurements are noted and taken manually (Jardemyr, 2013). This is very time consuming and it is demanding several visits when a measurement has been overlooked or a detail have not been documented (Jardemyr, 2013). Minor mistakes made while capturing the data lead to errors and simplified models of the reality.
The uncertainty and difficulty to verify new installations of new machines, gantry robots and equipment due to the lack of knowledge of the ceiling of the facilities often result in clashes and delayed projects or additional costs (Gustafsson, 2013). Having to move pillars, re-support the ceiling and rebuild walls just to get the equipment in place is not unusual (Gustafsson, 2013).
Clashes has also occurred in production when employees been unable to verify changes before implementation leading to costly stops in the production line (Bergman, 2013) and there is a wish from production engineers to be able to verify new products, at the development stage, in existing production lines and undocumented fixtures (Nordberg, 2013). This sort of request can also be found at the aftermarket method engineers who have problems validating if new service tools will fit inside the service stations, along with new truck models (Carlsson, et al., 2013).
1.1.2 Origin of the thesis
There is a need of new methods and tools in order to reduce stops, insecurity, waste and errors during the development and implementation of new factory layouts and products. During the last decade the tools and software supporting the industrial activities has developed rapidly such as CAD, CAM, digital assembly, virtual manufacturing and other simulation tools just to mention a few.
5
cancelled, mainly because the doubts of the usability of the results and how to manage the data (Kull, 2013; Rosengren, 2013). Scania CV AB started working with Digital Factory year 2009 (Hanson, 2013), a methodology that uses digital tools to examine and verify changes in the production or facility layout before the implementation in the real factory.
Through networking with other automotive companies, such as Volkswagen and Volvo Cars, Scania saw new possibilities in the laser scanning technology. During the last decade the technology of laser scanning has developed rapidly and the tools increased beyond just measuring and being a tool for facility services. By observing how Volvo Cars uses scanning in their business, today one of the world leading companies within the field of industrial 3D laser scanning, Scania renewed the interest of the technology and method which lead to this master thesis (Hanson, 2013).
1.2 Purpose
The purpose of this study was to investigate and evaluate the method and technology of 3D laser scanning in order to find benefits and drawbacks of using the technology within manufacturing industries. Furthermore, the study investigated how other companies use the technology today and if the methodology is profitable in a cost perspective. Another goal of the study was to deliver a process model of how a laser scanning can be preformed and how 3D laser scanning could be implemented in the current work flow regarding the factory layout development process.
1.3 Problem analysis
Based on the background and purpose four research questions were formulated in order to be answered in the thesis.
RQ1: How can the technology of laser scanning facilitate the development and maintenance of production systems within manufacturing industries?
RQ2: How does the automotive industry use laser scanning today? RQ3: How can laser scanning be profitable?
RQ4: How can laser scanning be implemented in the current layout work regarding the factory design process, at Scania?
1.4 Delimitations
Several delimitations were taken in consideration in this thesis.
The scanning performed at Scania was done by external consultants, based on good credentials and previous contact, using a Faro Focus 3D terrestrial laser scanner (TLS) and the associated registration software Faro Scene. Other consultants and equipment with associated software are available on the Swedish market but these have only been studied for comparative purposes. The experiences learned and the results from the study are based on the three scans preformed at the transmission machining, axle assembly and aftermarket service of Scania.
7
2 Methodology
This chapter is going to highlight the different research methods that were used during the study.
2.1 Method model of the study
This master thesis project was performed at the industrial development department, TEE, of Scania CV AB, one of the world’s leading manufacturers of heavy trucks and buses. Other important business areas are industrial and marine engines as well as services. The head quarter is locates in Södertälje, Sweden, and the production is located in Sweden, France, Netherlands, Argentina, Brazil, Poland, and Russia (Scania CV AB, 2012). In order to maintain competitive and profitable, Scania works continually to improve their processes and methods. The master thesis was considered as a pre-study of how the factory design process and general work of the engineer could be more efficient by using the technology of 3D laser scanning.
Due to the specific company focus, the design of the study was a case study. A case study design is useful when studying a process and possibilities of change (Davidsson & Patel, 2003), which was similar to the objective of this study. The case study can gain knowledge and great depth of understanding for the key company (Wallen, 1996) and it enables different methods of collecting data such as document study, interviews and observation (Yin, 2008). However, a case study is usually very time consuming and it might be difficult to finish, and it may risk to lose its general applicability if the case is too narrow (Wallen, 1996).
The current situation and previous research was studied both through primary and secondary data, i.e. data collected in the study via interviews and contacts, or information and data collected by someone else through articles and papers (Björklund & Paulsson, 2007). In addition to Scania, other companies currently using laser scanning was benchmarked in order to find the best practices from other industries (Boxwell, 1994).
Since the methods of finding output of the research questions differ depending on each question, they will be presented individually in the following sections.
2.1.1 Finding output to RQ1
To understand the technology and background of 3D laser scanning, information has been collected through articles, books, manuals, datasheets and interviews. In order to understand the current situation at Scania concerning needs, problems and thoughts several qualitative and semi-structured interviews together with seminars and demonstrations were held with production engineers, project engineers, project managers, scanning consultants and PhD students.
Two visits have been made to the Gothenburg based company ATS AB in order to learn how to use the two software Faro Scene and Bentley Pointools V8i for processing and editing point clouds, and thereby find and understand usable applications.
2.1.2 Finding output to RQ2
8
2.1.3 Finding output to RQ3
To get an indication whether the method of using laser scanning can be profitable and generate payoff within a couple of years a calculation was made based on collected information of expenditures and earnings related to scanning. The calculations were made with the help of the Scania MIKA-template, comparing the profit of the methods of internal and external scanning. However the MIKA-calculation should not be seen as a basis for investments.
2.1.4 Finding output to RQ4
To gain deeper knowledge of performing laser scanning and investigate how it can be implemented in the current layout process methodology, three Scania locations were scanned. The three locations were selected in order to include different business areas such as production and aftermarket. The production was divided into both machine processing and assembly. These three locations were considered as three different cases where it is possible to study how laser scanning can facilitate and support the work and processes related to each division.
The three cases will scan and study:
Machining group processing transmission gears, building 081 in Södertälje.
Assembly line of rear axles, building 210 in Södertälje.
Aftermarket service station, Kungens Kurva.
The observations and experiences gained from these activities provided a basis of involving and implementing laser scanning in the current factory design process methodology at Scania, using a Astrakan process model (Astrakan, 2010).
2.2 Data collection
Depending on the available resources and the purpose, background and length of the study, different methods can be used in order to collect data. According to Andersen (Andersen, 1994), there are three main ways: Document studies, Interviews and Observations.
2.2.1 Document study
The purpose of the document study is to use explicit data (Andersen, 1994) such as literature, annual reports and articles. The documents used in this study were mainly articles from journals, conference research reports, books, PhD dissertations, webpage information, supplier manuals, lecture slides and internal documentations at Scania.
2.2.2 Interviews
Interviews can be performed either verbally or through written surveys (Andersen, 1994). All interviews performed for this study were verbally with follow-up questions mostly answered through written correspondence.
Interview methods can be divided into structured, semi-structured and unstructured (Björklund & Paulsson, 2007). For these research interviews, the method of using semi-structured interviews were chosen in order to construct open-ended questions to allow discussion and spontaneous follow-up questions (Jacobsen, 1993; Björklund & Paulsen, 2007). With the main focus on the individual’s work tasks and reasoning about laser scanning, the choice of qualitative interviews was justified prior to quantitative (Trost, 2005).
2.2.3 Observation
9
can be observed directly and the researcher is always in direct contact to the objects being observed (Andersen, 1994).
11
3 Theoretical framework
The purpose of this chapter is to give a deeper understanding of the subject and current research. The chapter starts by briefly describing the terms of digital factory and digital manufacturing. Next section highlights and describes the current factory design process and a method of evaluating layout software. Further on, the chapter comprises a description of the background and technology of 3D laser scanners with the main focus on terrestrial laser scanning. Finally the chapter introduces the software used for editing and modeling of point clouds, Bentley Pointools V8i, with a description of its functions and claimed usability.
3.1 Digital factories & digital manufacturing
The digitization of production has introduced two concepts that are widely used in digital simulation and planning. These two concepts are digital factories and digital manufacturing. The modern production world often speaks about digital factories, but any general definitions are difficult to find. The terms are often vague and therefore companies and industries like to use and interpret the concept in their own way. The two terms are easily mixed up because they resemble each other, and they are sometimes incorrectly treated as synonyms, but their definitions and meanings are important to distinguish.
Digital factories concerns the technology used for capturing and representing information to model manufacturing systems and available processes in a factory (Kjellberg, 2006). The purpose of the digital factory is to mirror a factory and its available processes and therefore represent the relevant information of the factory’s resources and processes. These can be tools, machine tools, fixtures, conveyors, buffer and so on. The digital factory will also be a resource model that can be used as a base for preparation, plant design and layout of the production as well as being a tool for layout, material flows and analyzes (Kjellberg, 2006).
A digital factory is a model of a hypothetical or real manufacturing system, process or resource (Sivard, 2012). However, as a part of the digital factory, validation and optimization is done with digital manufacturing which should mirror the actual manufacturing through simulation and analysis. Digital manufacturing can therefore be defined as the technology used to process information in order to verify and optimize the manufacturing of products (Kjellberg, 2006).
The tools and utilities available within digital factories gain a lot of advantages. Although investing in new software and knowledge costs a lot, both financially and temporally, the profit is much larger in the long run. Building a brand new factory or only implement or install a new machine in an existing flow is very costly. Changes in a project are always more expensive at the end of a project than at the beginning. Making mistakes and having to redo in retrospect must be avoided and it is important to get it right from the start. Therefore, it is more cost effective to simulate the factory in a digital environment where collisions, laws and regulations can be checked and modifications and optimizations be done without affecting the existing production line in operation.
The digital way of working provides a streamlining of the entire work process including increasing product quality, reduced "time to market" while it enables interaction between supplier networks (Sivard, 2012). These systems provide an early verification and control of the process and a better opportunity to optimize and evaluate the factory before it is realized.
12
3.2 Factory layout
This section intends to describe what a factory layout is and what functions that should be included in factory layout software. It also describes the current factory design process, for developing and implementing factory layout, used at Scania.
3.2.1 Layout software evaluation
In a published article by Chen et al. (2012), the authors describe the layout process and what factors and functions that should be included and evaluated in layout software.
The objective of the layout process is to place the equipment in the best way possible so it enables an efficient material flow for the intended volumes and product families. The layout should visualize and confirm that the equipment fits inside a building and that the installations meets the required constraints regarding for example space, media (ventilation, electricity, and plumbing) and laws concerning safety and ergonomics. The layout can also be used to verify the environment for the operators working in it. The process of developing a layout can be divided into two phases; the concept phase and the detailed phase. In the concept phase, a block layout is created and used to show the product flow in the production. The layout typically shows which areas that are intended for operators, machines or conveyers and how the areas are supposed to be linked together. For example, the block layout can confirm that a machine’s material requirements can be satisfied from a nearby truck corridor.
In the detailed layout design phase, the physical connections and models of each system are evolving to become more detailed and exact. In this phase more detailed optimizations and simulations are done such as interdependencies between different equipment geometry. This can result in clashes which need to be resolved and redone in order to avoid and detect future problems. The detailed layout design is then ready to be implemented. But before the layout can be realized, the installation of the equipment and the constructions needs to be planned in detail to enable an efficient and safe implementation. The authors of the article (Chen, et al., 2012) has identified and summarized requirements and functionalities of a layout software based on industrial needs in order to enhance the effectiveness the layout process. The systems should be able to coordinate different layouts developed by different disciplines in a collaboration environment. It should also be possible to exchange and manage models and information from different sources. The requirements and functions, mostly related to creation, verification and modification of layouts, can be divided into five areas:
Creation of layouts models
Coordination of various models
Management of change and logistics
Verification of layout
Usability, efficiency and extendibility constrains
3.2.1.1 Creation of layouts models
The layout models describing the production can vary depending on the nature of the industry and what is manufactured. Generally the layouts include block layout, building layout, machinery layout, foundation layout and media layout (ventilation, heating, plumbing and electricity).
The layout can be represented in both 2D and 3D design or in a combination of them both. The 2D design might be considered as an old way of working but it still has some advantages and is used in many companies.
13
There is a well-developed archive of established drawing annunciations for measurements, doors and coordinates which mediates the information clearly, yet to be developed in 3D.
2D layouts are easy to print out which is an advantage if one lacks digital aids.
On the other hand the 3D design layout process is more agile, since it accelerates the layout process and provides the design team with rapid layout design and simulation. This can enhance the factory design and construction which will shorten the time to design a new layout or reconfigure an old one.
The 3D technique is an efficient tool to use when in consideration of more dimensions of the model, such as the height of a ceiling.
The 3D visualization makes verifications based on the human perception much simpler to detect, such as two components being too close to each other.
3.2.1.2 Coordination of various models
Models of buildings, media, machine tools and other equipment are often created by different stakeholders using different systems. One of the most important tasks in layout design is to consolidate and integrate the different models into the layout. In order to accommodate this, it is essential that the layout software can support several kinds of file formats from different system vendors without losing geometry, data or other information.
3.2.1.3 Management of change and logistics
Design and development of a layout is a gradual process. Small changes and updates are done continuously, resulting in many updates and different versions. At the same time many stakeholders may be using and modifying the same layouts and models, without even knowing it. Therefore, as one stakeholder modifies a corresponding layout or model, this will affect layouts designed by other disciplines and it is necessary that this change is managed correctly.
In order to facilitate this process it is preferred that the system can show the files current status and version number, e.g. draft, review, release etc., making sure that the right version of layout components from various sources is being used.
3.2.1.4 Verification of layout
A well created layout needs to prove and verify many various aspects according to safety requirements, legislations and specifications. The article (Chen, et al., 2012) mentions aspects and features such as:
Checking if the components will fit or collide either through viewing or automatic checking the geometry of the model. If using automatic clash detection it is preferable to be able to set tolerances of the collision.
Checking requirements and legislations according to safety and ergonomics based on if-condition rules.
Working conditions through immersion.
Checking product flow and productivity.
Walk through, which is an effective way of verifying a model and gain a rapid perception of the models potential.
The ability to set out and save notes in the model, allowing other users to easily read and gain understanding of the model, which can facilitate communications.
3.2.1.5 Usability, efficiency and extendibility constrains
14
Layout models are often very large when comprised by sub-models. A important functionality of layout software in order to increase the efficiency is the ability to simplify models by deciding the size range or scale of components that are desired to visualized. This saves computer power and disk space. However, there must be an agreement of level of detail to avoid mistakes and in order to be more extensible.
3.2.2 Factory Design Process
The current process regarding layout development and implementation at Scania, Appendix A, is divided into three major phases (Chen, 2010).
Pre-study
Project planning
Final planning and Installation
Each of the phases is divided to responsible units both internally and externally. The internal part refers to the production technique units at Scania. The external part consists of machine suppliers and DynaMate, who provide Scania with property and production maintenance and building projects.
3.2.2.1 Pre-study
When in need of changes in the production layout, e.g. a new machine, the process begins with a pre-study. This intends to lead to a concept and a decision if the project should be realized or not. The production engineers responsible of the specific area start with making a concept solution. This includes Value Stream Mapping, flow diagrams and a block layout. The block layout is based on LayCad models in DWG or DXF formats.
A request is then sent to DynaMate. They will evaluate the opportunities and the status of the current building such as investigating the foundations, the groundwater and if any explosive entries are needed to fulfill the requirements. The preconditions of the facility, evaluated by DynaMate, are then sent back to the production engineers who will create a rough layout. This rough layout should consist of construction and media documentations and shortly describe:
Number of machines
Rough electrical output
Human resources, the operator needs
Number of foundations
Number of heavy lifting
Break room / dining room / lounge with quantity
Number of office locations
Number of square meters
Process withdrawals
Cooling
Toilets
The rough layout should also be attached with a rough time table with stages and milestones. From this request DynaMate evaluate the consequences from the building and media measures. On this basis DynaMate delivers a cost proposal for the machine installation and the facility constructions. They also suggest improvement of the project and add comments to the timetable. With the input from DynaMate, the Scania production engineers receive calculation basis for the machine and facility investments.
15
3.2.2.2 Project planning
If the project is accepted and granted funding the pre-study enter the phase project planning. The production engineers create a requirement specification and send a request to machine suppliers. The machine suppliers then send back an offer and solution based on the requirements. Based on the different supplier solutions the production engineers develop different alternative layouts. The layout considered as the “best guess” is further developed into a preliminary detailed layout.
Mechanical installations, media and construction are added to the detailed layout which forms the basis for an updated requirement specification of the facility and results in a media layout. Based on the requirement specification of the facility DynaMate develop a detailed project plan. As a final step in the project planning phase the production engineers starts the procurement with the chosen supplier.
3.2.2.3 Final planning and Installation
After the procurement is complete the supplier starts building the new machine. DynaMate starts their own procurements with contractors and develop and provide construction documents for the reconstruction. With the input from the machine supplier and DynaMate the Scania production engineers enters the last phase. The layout is now “frozen”, i.e. considered as a final detailed layout which will not be reconfigurable. With the final planning complete the new machine can be delivered and installed according to the layout. As a last step in the final phase the layout is updated after the installation. This final layout is now considered as the current state.
3.3 3D laser scanning
Today laser scanning is applicable within many fields and laser scanners exist in many different shapes and sizes for different applications, although the basic technology often is the same. This section presents the history of laser scanning and technology with the main focus of terrestrial laser scanning, TLS.
3.3.1 History and background
Lasers in land and engineering surveying instruments has been widely used for the last 30 years as common parts in standard surveying instrument such as total stations, laser rangefinders, profilers, level and alignment devices (Shan & Toth, 2009). A natural development has been to add a scanning mechanism to a total station that were already equipped with laser rangefinders and angular encoders. Instead of measuring very specific individual points the laser scanner would allow the automated measurement and location of thousands of nonspecific points in the areas surrounding the position where the laser scanner instrument has been set up, resulting in a spherical point cloud with high accuracy in a few minutes (Shan & Toth, 2009).
According to Staiger (2011) the first laser scanners, as we know them today, appeared on the market in the mid to late 1990’s. Since then the market has seen dramatic improvements in terms of measurement speed, accuracy and general usability. At the same time all system became smaller, easier to handle and less expensive. Some of the first real applications was to scan dangerous environments with limited access such as nuclear power plants and offshore oil rigs and these industries were the main impetus for further development (Kull, 2013).
3.3.2 Terrestrial laser scanner
Terrestrial Laser Scanning (TLS) or Light Detection And Ranging (LiDAR) systems uses lasers to make measurements from a tripod or other stationary mount, a mobile surface vehicle, or an aircraft. The term LiDAR is sometimes used as a synonym to laser scanning but is more often associated with the airborne methods (Caltrans, 2011).
16
significantly for 3D measurements of structures. The technology has been used by civil engineer, landscape and architecture applications and projects for quite some time and a major reason for its success is that companies have realized that they, with the help of this technology, could have their facilities plotted in real size to a relative low cost (Bosché, 2012).
There are many benefits of working with laser scanning relatively traditional methods i.e. total stations, laser rangefinders or tape measure. These traditional methods usually takes time to do manually and they are often in need of proper planning to avoid future supplement. The scanner can output high resolute models of areas or objects that can be monitored easily. On the other hand the technique has some limitations. The quality of the models obtained from a scanner is relative to the file size of the data and it can take a significant time to complete a full scan. The systems collect a massive amount of raw data by determining the distance from the laser source and the horizontal and the vertical angles of the laser beam (Burton, 2007). A data point’s position can then be defined in space with a specific x, y and z coordinate. The point also receives a laser return intensity value and, if equipped with a digital camera, a RGB color code. The raw data product of a laser scan survey is then called a point cloud (Caltrans, 2011).
The scanner can only capture data from objects in front of it, so in order to re-create a complete 3D-model of an object, multiple scans has to be taken from different angles. The scans are then merged together to a joint point cloud.
3.3.3 TLS technology
Terrestrial laser scanners are classified in two different ways, the technique of measuring the distance and the type of beam deflection system (Staiger, 2011).
3.3.3.1 Distance measuring technique
The laser scanner uses a phase based method or a pulse based method of determine the distance to the object without an artificial reflector.
Time-of-flight systems use a focused pulse of laser light and wait for it to return to a sensor. The time it takes for the light to return, multiplied by the speed of light in air results in how far the pulse traveled. Since the pulse makes a round-trip, back and forth from the scanner to the object, the distance is divided by two (Curless, 1999). The accuracy of a time-of-flight system is therefore depending on how precise the scanner can measure the time since the light travel approximately 1 mm in 3,3 picoseconds.
In phase-based measurement technology, the laser scanner transmits an amplitude-modulated continuous-wave laser beam. The target distance is proportional to the phase difference and the continuous-wave length of the amplitude-modulated signal (Akin, et al., 2008). By using phase-shift algorithms the laser scanner determine the distance based on the unique properties of each individual phase by computing the phase difference between the emitted and reflected power signals (Curless, 1999), where the reflected power is provided by the amplitude of the reflected beam (Akin, et al., 2008).
17
Figure 1.Working principle of phase-based and time-of-flight (pulse-based) laser scanners (Akin, et al., 2008).
The pulse based system assures a wide measurement range but are slower than a phase based instrument. Vice versa, the phase based techniques allows a high measurement frequency but are limited in the range. Although, the difference between the two methods are becoming smaller with the latest generation of laser scanners (Staiger, 2011).
3.3.3.2 Type of beam deflection
Laser scanner can be divided into three types of different beam deflection system (Staiger, 2011). These types are camera-, hybrid- and panorama-scanners, which can be seen in Figure 2. The panorama type has the biggest Field-of-View (FOV) which is especially useful for indoor situations, and today the most common type on the market.
Figure 2. Classification of TLS by the type of beam deflection system (Staiger, 2011).
3.3.4 Quality of TLS
18
Table 1. The parameters that can affect the scan result according to Staiger (2005).
3.3.4.1 Materials, accuracy & range
The accuracy of the measurements captured by the scanner is dependent on the angle of incidence to the surface. Measurements taken from a surface that are perpendicular to the laser beam will produce better accuracies than those with a large angle of indigence to the surface. This may produce errors in the distance returned since the beam can elongate more with a larger angle(Caltrans, 2011).
Systematic test series have proven that different materials of the scanned object may have significant influences on the measurements of TLS, especially building materials such as concrete and metal plates (Voegtle & Wakaluk, 2009). The research performed by Voegtl and Wakaluk (2009) has shown that the range correction value increases with longer distances (nonlinear) and lower reflectivity of surface material. The standard deviation of range measurements also increased with longer distances, with about a factor of two, and reduced reflectivity. The intensity value were rather constant for different ranges but decreased with wider incidence angles and the standard deviation of intensity increased with higher reflectivity (Voegtle & Wakaluk, 2009).
A short comparison of the accuracy, working range and measurement rate of three of the most common laser scanners most suitable for industrial indoor application is listed in Table 2.
Table 2. Comparison of accuracy, range and measurement rate (Faro, 2013; Trimble, 2013; Leica Geosystems, 2013)
3.3.4.2 Referencing targets and merging scans
Objects or areas being scanned are often quite large and complex in shape. Therefore several different setups of the laser scanner should be made in order to capture the object completely from all necessary viewpoints (Reshetyuk, 2009). In order to obtain a complete representation of the scanned object or area, the scans should be transformed to a merged point cloud with a common coordinate system (Reshetyuk, 2009). Normally, the scanning is performed in such a way that the scans overlap pair wise. This means that a point cloud overlaps with the point cloud captured from the next scanner setup, see Figure 3.
Parameters affecting the scans
Object • Size • Curvature • Orientation • Surface Scanner • Performace of angels • Performance of distance • Calibartion • Synchronisation Enviroment • Vibration • Refraction • Optical Perturbations Method of data acquisition • Point density • Number and position of
reference points • Postion & number of
scans
• Distance from the object Method of calculation • Target Recognition • Registration • Calculation of elements
Brand Accuracy Range Measurement rate
FARO (Focus 3D) ± 2 mm 0.6 – 120m 976,000 points/sec
Trimble (Tx5) ± 2 mm (10-25m) 0.6 – 120m 976,000 points/sec
19
Figure 3.Illustration of the overlapping area when register to scans (Reshetyuk, 2009).
If single scans from different observation stations have to be merged together or transformed to a common coordinate system, it is possible to use some kind of referencing system of recognizable targets (Boehler & Marbs, 2002). It is desirable that these targets can be easily and accurately detectable by the scan registration software, such as spheres or plane target (Boehler & Marbs, 2002), Figure 4.
Figure 4. Spherical target.
20
Figure 5.Registration using targets (Reshetyuk, 2009).
To obtain sufficient information from the references and to fixate the scan in space the scanner needs to have at least 3 references in sight of every scanning position, Figure 6.
Figure 6. Example of scanning location setup, using references (Simplebim, 2013).
If necessary, it is also possible to use distinct natural point features, such as edges or corners of doors or windows, visible in the scanned point cloud to be registered (Reshetyuk, 2009). Although some of the accuracy of the merged scans might be lost in comparison with the spheres (Berlin, 2013).
21
Figure 7.External and scanner coordinate system (Reshetyuk, 2009).
ATS sells a kit with their own referencing system containing several spheres and the magnets. The magnets are possible to attach permanently by screwing them on for example concrete walls, pillars or steel objects. By having the magnet itself fixed it is possible to rescan an area of interest retrospectively and reuse the same reference positions. Since the references are at the same place as the old scan the new, updated, scan can be merged into to old point cloud with the same accuracy as the first scanning session (Berlin, 2013).
3.3.5 Bentley Pointools V8i
The Bentley Pointools V8i is a stand-alone pre-processing software for point clouds enabling 3D visualization and editing. The software is developed to be user-friendly and allow quick processing of point clouds and it is specially designed to handle larger point clouds, containing large amount of data (Bentley, 2013).
3.3.5.1 System requirements
To be able to use and run the software smoothly, the system hardware has to meet certain recommended requirements, Figure 8.
Figure 8. Bently Pointools V8i system requirements.
To fully support Bentley Pointools V8i functionality the graphics card must support OpenGL 2.0 or a later version. This should be supported on all ATI (AMD) and NVidia Graphics hardware since 2004 but may require a driver update. It is recommended to keep the installation of the software up to date by downloading the latest updates from Bentley’s website since they listen to the users and release regular updates and fixes.
Bentley pointools V8i
•Windows 8 or Windows 7
•Intel i7 or equivalent AMD quad-core processor, 2.6GHz or higher •8GB RAM
•NVidia or ATI (AMD) graphics card with 512MB on-board memory •200MB free disk space for installation
22
3.3.5.2 Using the software
Bentley Pointools uses its own native point cloud format POD (Point Database) that enables rapid background loading and compact file size through compression. A POD file does not require write-access, so it is possible to importing and displaying files from read-only media, read-only locations or shared files. The first time a POD file is loaded it will cache, which can take from few seconds to a few minutes depending on file size. After the file is cached subsequent loading will be instant.
The Graphical User Interface (GUI) has been designed to be as simple to use as possible and offers some ability to customize the layout to suit the user. The software uses a ribbon style toolbar that runs along the top of the window which makes finding the tools much faster by grouping families of tools together.
3.3.5.3 Import of different object types
Bentley Pointools V8i can import (or attach) the following types of object:
Point Clouds
3D Models
Drawings Point Clouds
A point cloud consists of a large number of points in space that describe an object. Each point in the cloud has a x, y, z coordinate and may also have additional properties such as color, reflective intensity or surface normal. The software can handle billions of points with modest hardware requirements since it is optimized for the display of point clouds.
The point cloud data can be imported from various file types from several laser scanner suppliers such as Faro, Topcon, Leica, Riegl, Optech, Trimble and Zoller-Fröhlich. However, the files are always saved in Bentley Pointools native file format POD to enable rapid background loading and compact file sizes. 3D Models
Textured 3D Models complete with material properties and transparency can be imported from a number of common model formats. However, there are currently a number of import and display limitations. Only UV, Planar and Box texture mapping modes are supported and NURBS or Sub patch / division surfaces are not supported.
Drawings
Most drawing primitives are supported. This includes lines, arcs, circles, text and dimensions. Bentley Pointools is able to import and display layered CAD drawing files and supports formats such as AutoCad DXF and DWG, and there is no need to explode blocks before import.
3.3.5.4 Project files, PTL
23
The results of point cloud editing operations are not saved back to the original POD files. This ensures that the workflow is non-destructive, since the point cloud usually represents captured data and this is usually the desired behavior in most industries.
3.3.5.5 Object tree
Objects in the workspace are managed in an object tree. All objects in the current project are represented such as overall point clouds information, individual point clouds with cloud information, notes database, drawing database with drawing layers, 3D Models database with 3D Model parts and 3D model materials. The object tree also gives the possibility to change the visibility of each object within the current project.
3.3.5.6 Navigation
The software enables two types of viewing projection; orthographic and perspective. It also enables two types of navigation; examine and explore. In Examine mode the viewer is rotated around the view target. In this mode the scene rotates relative to the camera. In the Explore mode the camera is rotated instead of the scene itself. This gives the feeling of first-person walking or flying through the environment.
The user is able to view the point cloud in RGB or intensity. Another possibility is to set the color ramp used to display the intensity values e.g. HUE. The user can save viewpoints, save rendered snapshots and as an extra effect view the point cloud in stereoscopic viewing (using 3D glasses).
3.3.5.7 Clipbox
Pointools features an editable clipbox tool. When viewing larger or complex point clouds it can be helpful to use this tool to isolate an area or volume of interest.
3.3.5.8 Notes
Pointools has the ability to store and display user created information in form of notes. A note can be attached to any point of a POD file, the end point of lines or corners of a 3D face and then be saved in the PTL project file. Once created, the note will be visible in the viewport and can contain text information and also a hyperlink to a web URL, a file, a saved view or an animation path. This function is ideal for presentations since it allows the user to attach documents, images, movies and other multimedia content to geographical locations in the point cloud.
3.3.5.9 Point editing
The software features a collection of tools for editing point cloud data for segmentation, clean up and color correction. A RGB painting tool let the user colorize or highlight areas or objects in the point cloud with any transparency. There is a set of tools that enables selection of part of the point cloud data were the selected area is highlighted. The selected highlighted areas can then be hidden from view using point visibility settings. This does not however delete the data or affect the original POD file, but enables selected areas to be isolated and exported as separate POD files. The selected point may also be moved between one of a possible 128 layers to isolate areas for detailed editing or for point cloud segmentation.
3.3.5.10 Layers
To enhance point cloud editing and workflow, Bentley Pointools V8i specializes the layer concept. All layer operations are saved in the PTL project file. Users of CAD software or design authoring software will probably be familiar with the general concept of layer based workflow.
3.3.5.11 Point transform
24
transformation, without really alter the values in the POD file. To make the changes permanent the file has to be re-exported as a new file.
3.3.5.12 Taking measurements
Measurements can be taken in two ways in the point cloud and the result is saved in the PTL project file. The measurement can either be taken by a single point position (point measure) which will show the selected points coordinate in the viewport. The other option is to a point to point measurement (distance measure) which will show the two points selected and the distance between them in the viewport.
The user will be able to:
Rename measurement
Add comments
Set the numerical precision of the shown output
Save the output measurements to a delimited ascii file which is a common format that can be read by most spreadsheets and databases.
Filter out coordinate values from the second point (dx, dy, dz). Useful to make simple measurements in only one direction.
3.3.5.13 Animation system
The Bentley Pointools V8i animation system is based on interpolation between selected keys placed on a timeline. The timeline is divided in frames but instead of setting up each frame it is possible to set up two or more key frames and the frames in-between are interpolated to produce a smooth motion. The technique is referred to as Key Framing. The basis of key frame animation is that the system interpolates between two keys to produce a smooth animation of the parameter value, which is any value that can be animated. In fact keys represent the value of a parameter at a particular frame.
The Graph Editor enables fine tuning of Keyframe position, value and interpolation method. The editor displays the changing value of the current selected parameter over the time as a graph with keys shown as nodes along the graph which can be adjusted. This means that it is possible to edit a objects coordinates relative a chosen camera position at a certain key frame on the timeline. It is also possible to parent one object to another object. The change of any parameters that belong to the parent will also affect the child object.
3.3.5.14 Clash detection and differencing
Bentley Pointools V8i has a toolset to test for clashes and differences. The clash detection will identify and report any interference between any two (or more) objects. The clash detection tools contain four types of test:
Static Interference
Dynamic Interference
Discrete Path Interference
Continuous Path Interference
This gives the ability to do one static test of an object, allow the user to move the object and see the clashes when they occur or test the clashes during a pre-set path at any given time or along the whole path.
25
3.3.6 Conversion of point cloud to CAD format
An existing method to convert point clouds to 3D models is through triangulation of the point cloud, where an approximated mesh between the points in the cloud creates a surface. However, no conversion procedures can recreate a surface from a point cloud flawlessly since the only information in the cloud is the points’ coordinates without any specific relation to each other and no guarantees exist of a point cloud completely free from disturbances (Mole & Araujo, 2010).
27
4 Results of case studies
This chapter will present the results of the case studies made during the master thesis. It will present benchmarking of another company’s strategies and the market of retailers and manufacturers, currently active in Sweden, with associated processing software. Next section will describe the locations and a time study from the case study scanning followed by usability testing of the two software Faro Scene and Bentley Pointools V8i. The chapter will also present cost calculations containing costs of laser scanning and a Scania payback calculation, MIKA. Finally this chapter will present the result of a methodology of working with scanning at Scania, Factory Scanning Process, and how it could be implemented in the Factory Design Process.
4.1 Benchmarking
To get an enhanced understanding of laser scanning in the automotive industry and the retailer and software market in general a benchmarking study are made. The automotive company that is studied is Volvo Cars. Similar to Scania, Volvo has their headquarters in Sweden and are at the cutting edge when it comes to using and developing laser scanning in the automotive industry. This background makes Volvo an interesting benchmarking case of how laser scanning is used in their business.
The laser scanning retail market in Sweden represents three brands of laser scanner equipment. These three are Faro, Trimble and Leica Geosystems. The benchmarking study is focusing on the retailers services and a comparison of the registration software provided for each laser scanner brand.
4.1.1 Laser scanning at Volvo Cars
According to Rönnäng (2013), Volvo defines virtual manufacturing as a technology and way of working to verify production systems virtually. This includes 2D and 3D software as well as PLM and laser scanning. With many years of experience, Rönnäng is one of the most qualified persons with knowledge and knowhow of laser scanning at Volvo.
Volvo Cars first came in contact with laser scanning somewhere between 1996-97. At the time, the focus was to scan single robot cells and compare the scans to layout drawings in order to verify the layouts. In 2009 they realized that the technology had matured and developed, mainly on the software side. The main reason of this new venture was that they could get a complete factory model from the combined scans in color and the software could handle larger point clouds, in the range up to 10.000 m2. This improved both
the visual scope and handling of the point cloud (Rönnäng, 2013).
Today Volvo Cars is one of the world leading companies of laser scanning in the vehicle industry both when it comes to operation and development. According to Rönnäng (2013), larger manufacturers such as Toyota, Mercedes and Ford are trying to implement scanning in their companies but they are facing difficulties and turns to Volvo for support and coaching. The others main focus is to scan robot cells but their purpose of scanning is rather different. As an example, Ford’s purpose is to save money in form of airline tickets and the availability of their engineers. Since they can analyze a production site remote through scanning, the engineers won’t have to spend valuable time and money on traveling.
Volvo Cars are continuously faced with a large amount of new and ongoing projects. In their business they have short time-to-market and they are working on several new car models simultaneously, produced in the same assembly line. Testing new car models were usually done with cardboard profiles being moved through the assembly line on Sundays when production is down and they have had problems with verification of new cars. Almost every new model being assembled have crashed in manufacturing line with the result of a scrap chassis, and worst of all, stop in production since they assemble all different models in the same line (Rönnäng, 2013).
28
discussion base for communication and a tool for simulation and verification of products and processes digitally. Continued efforts were motivated by strategic investments of the company’s methodology processes. Today’s users are mostly simulation- and planning engineers and maintenance engineers. There are currently about 50 employees using the laser scan result in their daily work and 3 are working with development. However, they estimate that about 300 would benefit from working with the results and a vision is that everybody would have access to and a majority be working with the scanning results in 2017. Volvo Cars are currently using several different software, such as:
Bentley Pointools V8i
Faro Scene
Faro Web-share
IPS (Industrial Path Solutions)
Process Simulate (Siemens PLM)
The actual scanning is performed by an external consultant company, ATS AB in Gothenburg. Volvo estimates that they will run the scanning internally by themselves somewhere in 2014, focusing on local rescans. The major global update, when rescanning the complete factory, approximately every 2-3 years, will however probably be scanned by consultants because of the time consumption. The scans are performed at the functional areas, covering the lines and the logistics. About 60% of their factories have been scanned, the final assembly remains and they have just started in their new factories in China.
The laser scans produce a lot of data, in the order of Terra Byte. Large data files generates loading times and rescans may create information conflicts. Today, Volvo Cars handles the massive amount of data with a file structure at their intranet. Although, they have a dialogue with Siemens about a PLM system that supports the management of scan data and files. When scanning, Volvo takes the factory and building origin in consideration which they have also measured with a total station for higher accuracy, since when placing a new machine, the machine origin is as important as the buildings. When using measured reference points form the total station the experienced final scan results have accuracies around ± 3 mm, sometimes better.
Volvo doesn’t believe that scanning will replace 2D layouts completely. They will continue to scan the existing business and use CAD when building something new. There has been great interest of the technology within the company, especially when showing the visual results. However, there are problems finding owners of the technology and the data produced.
Rönnäng (2013) states that the point cloud can be used to verify the layouts according to the reality, which would decrease the risk of making wrong decisions based on out of date layouts or distorted perception of reality.
Positive experiences
o 3D model of reality
o Great visual tool for communication o Verify layouts
o Simulation and validation of new products
o Decreased travel costs – higher engineer availability o Generally positive reception of the technology
Negative experiences
