Cable harness design, assembly and installation planning using immersive virtual reality

Download Cable harness design, assembly and installation planning using immersive virtual reality

Post on 15-Jul-2016




0 download



    Cable harness design, assembly and installation planningusing immersive virtual reality

    James M. Ritchie Graham Robinson Philip N. Day Richard G. Dewar Raymond C. W. Sung John E. L. Simmons

    Received: 1 August 2006 / Accepted: 1 March 2007 / Published online: 3 May 2007

    Springer-Verlag London Limited 2007

    Abstract Earlier research work using immersive virtual

    reality (VR) in the domain of cable harness design has

    shown conclusively that this technology had provided

    substantial productivity gains over traditional computer-

    aided design (CAD) systems. The follow-on work in this

    paper was aimed at understanding the degree to which

    various aspects of the immersive VR system were con-

    tributing to these benefits and how engineering design and

    planning processes could be analysed in detail as they are

    being carried out; the nature of this technology being such

    that the users activities can be non-intrusively monitored

    and logged without interrupting a creative design process

    or a manufacturing planning task. This current research

    involved the creation of a more robust CAD-equivalent VR

    system for cable harness routing design, harness assembly

    and installation planning which could be functionally

    evaluated using a set of creative design-task experiments to

    provide detail about the system and users performance. A

    design task categorisation scheme was developed which

    allowed both a general and detailed breakdown of the

    design engineers cable harness design process and asso-

    ciated activities. This showed that substantial amounts of

    time were spend by the designer in navigation (41%),

    sequence breaks (28%) and carrying out design-related

    activities (27%). The subsequent statistical analysis of the

    data also allowed cause and effect relationships between

    categories to be examined and showed statistically signif-

    icant results in harness design, harness design modification

    and menu/model interaction. This insight demonstrated that

    poorly designed interfaces can have adverse affects on the

    productivity of the designer and that 3D direct manipula-

    tion interfaces have advantages. Indeed, the categorisation

    scheme provided a valuable tool for understanding design

    behaviour and could be used for comparing different

    design platforms as well as examining other aspects of the

    design function, such as the acquisition of design decision

    intent. The system also demonstrated the successful auto-

    matic generation of cable harness assembly and cable

    harness installation plans from non-intrusive user-system

    interaction logging, which further demonstrates the

    potential for concurrent design and manufacturing planning

    to be carried out.

    1 Introduction

    The use of interactive immersive virtual reality (VR) will

    become prevalent in a number of forms over the next few

    years within the product design environment. The appli-

    cation of this technology will probably mirror how

    expensive turn-key computer aided design (CAD) systems

    began to impact on industry in the late 1970s and eventu-

    J. M. Ritchie (&) G. Robinson P. N. Day R. G. Dewar R. C. W. Sung J. E. L. SimmonsScottish Manufacturing Institute, Heriot-Watt University,

    Edinburgh EH14 4AS, UK


    G. Robinson


    P. N. Day


    R. G. Dewar


    R. C. W. Sung


    J. E. L. Simmons



    Virtual Reality (2007) 11:261273

    DOI 10.1007/s10055-007-0073-7

  • ally became generally available on low-cost PC-based

    platforms with extensive real-time solid modelling capa-

    bilities. In the recent past the focus of immersive VR

    applications has been mainly in the research laboratory and

    in larger companies; however, as this technology becomes

    cost effective and more widely used in the design and

    manufacturing engineering sector, it is important to

    understand how to analyse its use and evaluate its benefits

    and limitations as it begins to impact on creative engi-

    neering processes such as conceptual and detail design and

    manufacturing/assembly process planning.

    In this paper the main focus is on using head-mounted

    display (HMD) immersive VR as a tool for the analysis of a

    creative design task, i.e. the 3D generation of cable harness

    routes. We also investigate how the engineering designer

    approaches a problem and what the key issues are with

    regard to future virtual design systems of this kind. Spe-

    cifically, a task categorisation of system usage is outlined

    and tested for analysing user activity in a computer-based

    cable harness design application.

    This paper initially focuses on immersive VR as an

    enabling technology in engineering design and then

    emphasises the specific problems and solutions associated

    within the domain of cable harness routing. It then

    describes the immersive VR apparatus and experimental

    methodology used to investigate how virtual engineering

    design tasks are carried out for cable harnesses followed by

    a detailed analysis and discussion of results. This is

    followed by a section demonstrating the potential for the

    automatic generation of downstream manufacturing plan-

    ning data from user activity logging before drawing some


    1.1 Immersive virtual reality

    Virtual reality itself takes many forms with a wide array of

    technologies classified as being virtual environments (VEs)

    of one form or another. There are now many applications

    where VR is being used for mechanically engineered

    products and a wide variety of different types of technol-

    ogy that can be applied in engineering domains (Jayaram

    et al. 2001). This paper focuses on the HMD, where the

    user is surrounded by a virtual world generated by com-

    puter graphics; the models within this can be interacted

    within real time (Fig. 1) depending on the input devices

    and tracking devices attached to the system. The helmet

    incorporates sensors to track the users physical move-

    ments as well as allowing for relative sound input.

    Therefore, if HMDs are to be used within the design

    platforms of the future it is necessary to carry out research

    to determine how system interfaces need to be designed to

    enable this technology to be used to its full effect. Cur-

    rently, HMD research has shown that there are health and

    safety issues to address, such as heterophoria change, vir-

    tual simulation sickness and oculomotor problems. These

    effects must be understood to support the future develop-

    ment of product engineering design systems using this kind

    of design platform leading to, for instance, the recom-

    mendation for maximum length exposure durations of

    approximately 20 min (Health and Safety Executive 2000;

    Howarth 1997, 1999; Howarth and Costello 1996, 1997);

    therefore, the cable harness design work tasked in this re-

    search was set at this time limit.

    1.2 Typical virtual engineering applications

    In the area of product design, VR systems can change the

    way in which engineers develop products and work

    together to generate ideas, embody concepts and produce

    the information necessary for cost-effective manufacture

    (Jayaram et al. 2001). Gomes de Sa and Zachmann (1999)

    see a role for immersive VR throughout the whole product

    development cycle, outlining the extensive application of

    this technology for digital mock-ups. They feel that im-

    mersive VR, must be at least as easy as designing witha CAD system. Ng et al. (2000) discovered that this was

    the case because the training times for using an immersive

    VR design system were much shorter than that associated

    with traditional CAD systems. Cruz-Neira et al. (1992)

    used a C2 CAVETM environment for architectural design

    so that students could appreciate a model at full scale. They

    mention the importance of recording and storing some

    form of design intent as activities are carried out. Weyrich

    and Drews (1999) used a virtual workbench to design and

    found that the method appears to effectively support how

    engineers think during the design process, which points to

    the importance of collecting subjective information relating

    to how designers feel whilst participating in a creative

    activity supported by virtual technologies. These findings

    Fig. 1 Head-mounted display (HMD)

    262 Virtual Reality (2007) 11:261273


  • also highlight the need to be able to breakdown design

    processes to determine how systems development can take

    place to support product engineering activities. ChiCheng

    et al. (2002) developed a series of interface tests and a

    classification of interaction activities to investigate how

    designers interact with an immersive virtual product design

    studio during a tightly constrained set of design tasks. Their

    paper does not elaborate as to what form these take but

    does show that VR gives advantages over the traditional

    interaction approach. However, these tasks required little

    original user input and deliberately neglected thinking


    With regard to designers thinking, this highlights the

    importance of cognitive issues in design. Design is a cre-

    ative act, described by McPhee (1997) as a mysterious mix

    of science and art that can only be understood by first

    understanding how humans think and behave. He also

    suggests design is instinctive; a notion echoed by Schonss

    (1991) knowing-in-action. Theory also proposes that

    where instinctive activity does not lead to a satisfactory

    outcome, the designer suffers a breakdown, a difficulty

    that makes tacit reasoning more explicit (Guindon et al

    1987). Furthermore, studies repeatedly show design to be

    unsystematic and ad hoc at the level of an individuals

    actions, despite the influence of an explicit rationalistic

    guiding procedure (Cross 2001). Even when using the same

    methods, it has been noted for some time that designers

    produce appreciably different designs (Adelson and Sol-

    way 1985). It is further suggested that a designers

    behaviour may be a function of the designers cognitive

    load (Guindon 1990) and, as a consequence of this, the

    related notion of modal shifts also emerges (Akin and

    Lin 1995). Here, the designer is in a particularly creative

    state (as indicated by important, novel decisions being

    made) and rapidly alternates between tasks in recognisable

    patterns. The nature of the immersive VR platform of the

    kind used in this research provides the potential for the

    non-intrusive analysis of design tasks and the recognition

    of these associated patterns in a manner which would be

    very difficult with traditional CAD systems. This is

    potentially further amplified in downstream manufacturing

    planning task extraction in the process design phase of

    product development.

    The capability of VR during creative design tasks was

    demonstrated by COVIRDS (COnceptual VIRtual Design

    System) which showed the interactive capabilities of im-

    mersive virtual design (Dani and Gadh 1997) using hand

    tracking and voice input for a VR-based CAD environ-

    ment; allowing rapid concepts to be modelled through free-

    form shape creation. Varga et al. (2004) have also inves-

    tigated the use of hand motion as a means of creating

    conceptualised geometry for design purposes and suggest a

    novel classification scheme for categorising these motions,

    i.e. contact, speed, adaptability and fidelity; focussing more

    on free-form geometry than more precise point-to-point

    sketching scenarios. Work at Heriot-Watt University (Ng

    et al. 2000; Holt et al 2004) showed that immersive VR

    also has a role to play in the design process.

    As can be seen from this review, there is still a need to

    understand how the creative design process can be analy-

    sed in detail when applying HMD VR technology to the

    design task, how these activities are broken down and

    where the emphasis on interface and technology develop-

    ment should take place to further advance the technology.

    1.3 Research work domain

    Cable harness design has been a classic design problem for

    many years because even with the application of extensive

    CAD-based packages available for this task many compa-

    nies still employ physical prototypes for the generation and

    checking of cable routes (Ng et al. 2000). Early cable

    harness design work was carried out in the USA in the

    1990s in an attempt to automate the choice of a cable

    harness route (Conru 1993), with subsequent work using

    genetic algorithms to tackle the same problem (Conru

    1994). Wolter and Kroll (1996) routed strings around

    solid parts, and in some projects robot path planning was

    applied to piping systems as a routing solution (Zhu and

    Latombe 1991). Work at Heriot-Watt University (Ng et al.

    2000) showed that immersive VR has a role to play in this

    design process, and research at Iowa State University

    (Fischer et al. 2002) employed a VR system for routing

    flexible hoses that validated VR as a practical tool but did

    not analyse its effectiveness as an interactive design tool.

    Early work at Boeing (Caudell and Mizell 1992) in the area

    of augmented reality indicated the advantages of virtual

    technologies in assembling cable harnesses.

    A survey of industrial companies showed that there was

    a need for human expert intervention to make fine adjust-

    ments and verify solutions (Ng et al. 2000); therefore it is

    timely to investigate the nature of new human-driven tools

    to support interaction with data in this domain. The key

    issue is the integration of the human expert into the sys-

    tem by treating the operator as an integral part (Holt and

    Russell 1999). This approach emphasises the need to

    examine creative design activities in more detail to see how

    tools and methods can be introduced to support the cable

    design task. The efficient and reliable manufacture of

    cabling systems for many products in such sectors provides

    designers with a range of challenges. Cable layouts are

    often so complex that design tends to be carried out as an

    end activity, which may lead to higher costs or even a

    product redesign. The problems encountered during the

    cable harness design stage have a marked impact on the

    time needed for new product introductions with multiple

    Virtual Reality (2007) 11:261273 263


  • revisions of physical prototypes being commonplace (Ng

    et al. 2000).

    VRs unique capability to immerse the user in a de-

    sign experience makes it a useful domain in which to

    carry out detailed design studies, whilst cable harness

    design represents a convenient ring-fenced design task

    which can be measured and analysed in isolation; how-

    ever, it is flexible enough to allow some form of task

    variety to be built into system experiments in a con-

    strained design environment.

    Earlier work at Heriot-Watt University in the area of

    cable harness design compared an immersive VR design

    environment called CHIVE (Cable Harnessing in Virtual

    Environments) with a number of CAD systems and

    demonstrated that HMD virtual technology gives pro-

    ductivity benefits during creative cable routing design

    activities. This showed conclusively that VR provided

    substantial productivity gains over traditional CAD sys-

    tems (Ng 1999). The follow-on work discussed in this

    paper was aimed at understanding the degree to which

    various aspects of the immersive VR system were con-

    tributing to these benefits and how engineering design and

    planning processes could be analysed in detail as they are

    being carried out. The nature of this technology is such

    that the users activities can be non-intrusively monitored

    and logged without interrupting a creative design process

    or manufacturing planning task providing considerable

    potential for understanding creative design activities with

    no interruptions to cognitive thought processes. Central to

    this research was the use of a more robust, CAD-equiv-

    alent VR system for cable harness routing design, harness

    assembly and installation planning which could be func-

    tionally evaluated using a set of creative design-task

    experiments to provide detail about the system and users

    performance. This was based on the table-top metaphor

    (see Fig. 2) using comprehensive user logging and was

    developed to non-intrusively collect detailed information

    relating to the design solutions and approaches used by a

    number of engineers, as well as automatically generate

    assembly plans from user interactions. It was decided that

    the majority of the design tasks undertaken for the initial

    experiments would focus on the 3D volumetric design

    process because this was considered by the companies to

    be a priority with regard to cable route planning as 2D

    schematic design could be easily handled using proprie-

    tary packages.

    2 Apparatus and methodology

    2.1 Apparatus: COSTAR experimental platform

    The system developed as an experimental platform for this

    research was called COSTAR (Cable Organisation System

    Through Alternative Reality). This was implemented on an

    SGI Octane2TM with V12 dual head graphics drivingeach eye on a V8 stereo HMD. Peripherals attached to the

    system include a Flock of Birds magnetic tracking sys-tem and Pinch Gloves. The software platform used fordeveloping the COSTAR system was the SENSE8WorldToolKit release 9.

    The COSTAR system enables the engineer to design and

    assembly-plan cable harness assemblies within the im-

    mersive VR environment, with all design functions,

    including the creation of new objects, being performed

    while they are immersed in the system (Fig. 3). Interac-

    tions with the system are achieved by means of a custom-

    built menu system and pinch gestures, with combinations

    of two to ten touching fingers, in addition to the spatial

    input afforded by the Flock of Birds system.

    Fig. 2 Workbench metaphor,from (Holt et al. 2004)

    264 Virtual Reality (2007) 11:261273


  • As the prototype system is fully immersive using two

    gloves and a HMD, menus had to be designed for ease of

    use. The current system uses a hierarchical 3D (ring), as

    applied by Liang and Green (1994), and more recently by

    Gerber and Bechmann (2004; Fig. 4). The engineer can

    input the cable harness routes by plotting points in 3D

    space, these being joined together to produce the cable path

    itself. Subsequent editing of the cables is possible by

    selecting the plotted points and bending them around

    obstructions, bunching or pulling them together to form

    cable bundles, inserting additional points and adding con-

    nectors and fasteners; depending on the menu options

    chosen. Figures 5, 6, 7 and 8 show the system with various

    operations being performed.

    COSTAR logs all of the users cable harness design and

    assembly activity-related actions, with the position of the

    hands and head being logged approximately 50 times per


    3 Experimental procedure

    As mentioned earlier, the experimental cable harness design

    tasks were to be completed in around 20 min for health and

    safety reasons. Three loosely constrained creative design

    tasks were organised to evaluate the utilisation of each

    designers time. The tasks covered the common design

    activities for cable harness processes, such as routing,

    bundling, cable modification and choosing connectors. The

    log files from these activities were subsequently decom-

    Fig. 3 The COSTAR cable harness design system

    Fig. 4 Hierarchical ring menu

    Fig. 5 Creating a cable from point to point

    Fig. 6 Inserting a cable point

    Fig. 7 Model on completion of the experimental tasks

    Virtual Reality (2007) 11:261273 265


  • posed and analysed in order to ascertain the areas of the

    virtual cable harness design system that were used, the

    kinds of activity the designers performed and their distri-

    bution throughout the total design time taken. Since this was

    a detailed design study there was a need to provide the

    participant with a realistic design problem for which they

    then had to provide a solution; the major goal being to

    evaluate the ways in which the system and technology

    supported or hindered the engineer during their work, and

    how the engineer tackled the design problem itself. There-

    fore, participants were given sufficient information about

    what the goals of the task were along with its main

    boundary conditions but were then free to determine what

    form the final design solution should take. This uncertainty

    of task outcome prevented the evaluation process becoming

    a prescriptive controlled experimentthe intention being to

    give participants a sense of doing a real design activity with

    the system. This reflects our earlier discussion of design

    itself being a creative act that could not be assessed by a

    rigid process. All of the tasks were associated with typical

    cable harness design practices within the industrial partners,

    were carried out within the same product model (Fig. 7)

    and involved consecutive stages of the overall cable harness

    design process; namely (1) outline design; (2) detailed de-

    sign; and (3) redesign. Tasks 1 and 2 were used mainly for

    participant training and familiarisation, and task 3 for de-

    sign task analysis.

    (1) Outline design: The first task was to generate two new

    electrical interconnections within the product model.

    Each of these interconnections had to join two spe-

    cific connectors within the model and have a specified

    cable type. The goal of this task was to define the

    electrical interconnections that would be provided by

    the harness rather than to produce a representation of

    the physical harness design, and hence, the routes

    followed by the cables were not important.

    (2) Detailed design: The second task contained pre-de-

    fined cable interconnections in a model, a number of

    which had already been routed through a sequence of

    cable clips to produce a harness design. It also had

    three other cables that defined electrical intercon-

    nectivity but had not yet been routed to produce a

    physical path for these cables to follow within the

    harness assembly. The user was instructed to, route

    the outline cables in the model through the cable clips

    to complete the cable harness design. However, the

    individual participants needed to use their engineer-

    ing judgement as to what the completed harness

    design should be and how to achieve that goal. Fig-

    ure 8 shows a partially completed route.

    (3) Redesign: The third and final task started with a

    product model that contained the design of a com-

    pleted harness assembly. The participant was then

    given some engineering change requests requiring

    redesign of the harness in some manner. The specific

    changes required were the addition of a new cable to

    the harness and the removal of one of the cables and

    its associated connectors. Finally, there was another

    undefined error within the model that the partici-

    pants were required to locate and fix. This undefined

    error was a cable being routed through a solid wall,

    with the cable therefore requiring re-routing.

    Ten participants completed the experiments, out of

    which nine were drawn from the engineering staff and

    student populations of the university and the tenth being an

    engineer drawn from industry. All the participants were

    male, eight were 2029 years of age and two were 30

    39 years of age, all with normal or corrected-to-normal

    vision. Everyone was right-hand dominant with eight being

    right eye and two being left-eye dominant. Seven of the

    participants estimated that they had between 10 and 100 h

    of previous CAD experience with three estimating 100

    1,000 h experience. Seven also had no prior VR experi-

    ence, two had less than 10 h and one had 100-1,000 h of

    VR exposure. Identical session structures were used at each

    of the three evaluation task sessions. The immersive design

    activity was followed by a semi-structured interview dur-

    ing which feedback about the system, and the participants

    experience with it, was collected.

    4 Analysis of results

    Data collected via log files included performance and usage

    data. In addition, post-experiment data was collected in the

    form of system usability and functionality data, along with

    informal subjective discussions regarding system perfor-

    mance and future changes. From the results for task 3, the

    Fig. 8 A partially completed route

    266 Virtual Reality (2007) 11:261273


  • usage of the system was analysed by means of various

    novel categories of functionality and system state that were

    developed for these experiments and followed on from the

    broader categories applied by ChiCheng et al. (2002). The

    new categorisation reflected general system usage and

    allowed the analysis of key parameters and functions dur-

    ing a users interaction with a computer-based design tool,

    such as CAD or VR. In relation to the working environ-

    ment, it was decided to analyse the time spent in the model,

    in help screens and in the menus so that their influence on a

    design task could be compared against a proposed detailed

    activity task categorisation.

    As a result, we developed environmental categories and,

    on analysing the log files, produced the distribution shown

    in Table 1 and Fig. 9. These data demonstrate the average

    percentage of time spent in each of the new environmental

    category subdivisions as the designers completed design

    task 3. We can see that a high percentage of the time (69%)

    involved users carrying out activities within the model and,

    to some extent, being creative. Only a small proportion of

    the time (8%) was spent in help/task instruction, sup-

    porting the informal feedback from the users that the sys-

    tem was easy and intuitive to use.

    After analysing the data and the associated design

    process activities, the various action sequences within the

    log files were grouped together to enable a numerical and

    statistical analysis of the cable harness design approach

    used by the participants. In the lexis of cognitive task

    analysis, these action sequences are often called task

    plans (Preece et al 1994). From these data, a set of

    design activity categories were defined so that participant

    activities could be compared and correlated between each

    other and the environmental categories in Table 1. The

    four action sequences, or activity categories, chosen


    (a) Design: all activity that the user carries out to directly

    amend the design solution or associated documenta-


    (b) Information: all user activity which involves them

    acquiring information from a text screen.

    (c) System operation: all activities which are required by

    the user to operate the system but does not affect the

    design solution.

    (d) Navigation: all activity which modifies the partici-

    pants viewpoint of the model but does not normally

    change the design solution itself.

    However, due to the fact that design was at the core of

    CO-STAR system this was further subdivided into three

    subcategories to allow more detail to be obtained regarding

    an analysis of activities carried out whilst the user was

    being creative during the design task. These subcategories


    (a) Designgoal: user actions which alter the design

    solution/model and advance the design towards its

    final state.

    (b) Designsupport: activities which do not produce a

    change to the design solution but enable the user to

    subsequently alter the design.

    (c) Drag and drop (position edit): the movement of an

    object by the user interactively within the model


    The results from this categorisation structure are shown

    in Table 2 and Fig. 10.

    From these results it can be seen that a large proportion

    of the time was spent navigating around the model (41%).

    This reflects the experimental model being presented to the

    designer in super-scale, i.e. the model surrounded the

    engineer. Flying was employed as the navigation mode

    because it was the traditional type of navigation used with

    an HMD. The flying speed was kept constant in order to

    reduce confounding variables in experimentation. How-

    ever, with such a large proportion of the time being spent

    moving around the model, the categorisation scheme shows

    that, during the creative design process, it would be

    advantageous to reduce navigation time considerably. This

    backed up an important finding from Ng (1999), which

    found that user scaling of the virtual model while immersed

    considerably enhanced the designers perception of the

    product model and the associated design task, as well as

    reducing the quantity of navigation necessary.

    Table 1 Environmental category subdivisions for design task 3

    Environmental categories

    Model Help/task


    Menu (no

    model visible)





    Mean time (s) 867 101 289 0 1,257

    Mean time (%) 69 8 23 0 100





    Menu (No modelvisible

    Fig. 9 Average time distribution for environmental category subdi-visions

    Virtual Reality (2007) 11:261273 267


  • Furthermore, a large number of sequence breaks or

    pauses were apparent during the cable routing process,

    implying that idle time potentially exists in the process.

    Many of these breaks existed within the navigation of the

    model, which points to the need for more effective navi-

    gation tools as well as an analysis of what is happening

    when these breaks are taking place. In addition, this time

    could indicate areas in the process where the designer is

    thinking about the design. This prompts the research

    question, Can thinking time be identified and analysed in

    some way in order to evaluate design intent?

    Around 27% of the time within the system was spent on

    design-related tasks. Although a high percentage of the

    time, it was apparent from our research that there were

    opportunities to improve the interface in terms of naviga-

    tion and menu interfaces to free up even more time for

    creative design. In this paper, we deliberately do not report

    on the findings and recommendations prompted by our

    usability and system functionality data for brevity and to

    maintain a focus on activity categorisation issues.

    Furthermore, it was important to understand how much

    time was being spent by designers in unproductive activi-

    ties and when there were breaks in the actual process of

    interfacing with the system, whether in design, menu

    operation or navigation. In order to do this two supple-

    mentary categories were developed which were generic

    across all of the design categories, namely:

    (a) Unproductive activity: all category activity that can be

    removed from the process without affecting the final

    outcome of a task.

    (b) Sequence breaks: abeyance in activity between the

    end of one action sequence and another with no input

    from the user. For instance, user thinking time, or an

    activity that did not register as an interaction, such as

    a head or a hand movement.

    The results of these further subdivisions applied across

    all of the existing categories and across all of the experi-

    mental tasks are shown below in Table 3.

    What these data highlight is that there are substantial

    parts of the process during which the users are taking

    breaks from carrying out any form of activity (28%). Al-

    though this time may illustrate when they are simply

    resting, it is apparent that with the task duration being so

    short this time could be associated with thinking time

    about, for instance, menu interfacing, design, design

    modifications, etc. These issues will require further inves-

    tigation; however, this analysis shows that utilising a de-

    sign categorisation scheme and having the ability to carry

    out the detailed monitoring of activity in a computer-aided

    engineering environment could potentially provide a means

    of non-intrusively analysing design intent.

    Another major outcome of having the ability to carry

    out the detailed analysis and categorisation of a design

    process in this way is the ability to investigate statistically

    the cause and effect relationships between the various

    categories and subcategories. Consequently, tests were

    carried out to see if any significant relationships could be

    identified. This analysis compared all of the environ-

    mental categories and activity categories together. Using

    the ShapiroWilks normality check test of Goodness-of-

    Fit (W) the data were found not to be normally distributed

    (p > 0.1). This would be expected in an open-ended

    Table 2 Time distribution foractivity categorisation

    Activity category Totals





    Drag and


    Information System



    Mean time


    131 57 157 106 296 510 1,257

    Std. dev. 52 13 130 36 77 160

    Mean time


    10 5 12 8 24 41 100







    Design GoalDesign SupportDrag&Drop



    Fig. 10 Average time in activity categorisations

    Table 3 Supplementary category subdivisions

    Activity category

    Unproductive activity Sequence breaks

    Mean time (s) 70 356

    Std. dev. 51 110

    Mean time (%) 6 28

    268 Virtual Reality (2007) 11:261273


  • creative design task of this kind where considerable

    freedom of expression was given to the engineers to

    generate a final solution to the associated cable harness

    design problem. Because of this finding, non-parametric

    correlations were evaluated by means of a Spearmans

    Rho test. The following significant, and tending-towards-

    significant, correlations were found, as summarised in

    Table 4, where significant was defined as p < 0.05 and

    tending-towards-significant was defined as p < 0.1. The

    majority of comparisons showed no significance and have

    been omitted here for brevity.

    The results illustrate some obvious and not-so-obvious

    cause and effect relationships between the various activity

    and environment categories used within the experimenta-

    tion and give an interesting and novel insight into the cable

    harness design process itself as well as the functionality of

    the immersive virtual reality design system.

    If engineers spend less time in the model being creative

    then they usually spent potentially useful time in design

    support (Pair 1: q = 0.89, p < 0.05). Although this mightinitially appear to be an obvious statement, this result

    validates the approach and categorisation comparative

    structure used because it reflects the actual system usage

    observed and reflects opinions expressed by designers in

    the post-experimental interviews. Designers also tend to

    access menus and help/task information in design support

    (Pair 2: q = 0.69, p < 0.05; Pair 3: q = 0.85, p < 0.05)which shows the usefulness of the categorisations applied

    for analysing a cable harness design task. This stresses the

    need to design menu, help and support information and

    their associated interfaces as efficiently as possible to

    maximise creative design time and to minimise menu

    interaction and support tasks; something which was

    apparent from the basic categorisation analysis but is

    strongly supported by the statistical comparison. Less time

    in the menus means more time in the model, i.e. carrying

    out productive design (Pair 4: q = 0.89, p < 0.05). Sim-ilarly, less time in help means more time doing productive

    design (Pair 5: q = 0.90, p < 0.05); two strong negativecorrelations and, again, perhaps obvious observations;

    however, the relationship between these is now quantifiable

    using the approach developed in this work.

    Continuing through Table 4, when the designers spend

    more time carrying out unproductive activities, they were

    most likely to be in the menu environment than doing other

    activity (Pair 6: q = 0.69, p < 0.05), which implies thatusers are at their most productive in the modelling (design)

    environment. This supports the categorisation scheme

    developed because it numerically confirms a previously

    subjective judgement when observing such a design task.

    Tending towards significance, drag-and-drop is also

    shown to be an important productive design activity (Pair

    7: q = 0.61, p < 0.1) because the designer is improvingthe design and justifies the categorisation because it iden-

    tifies the drag-and-drop task as important to the design

    activity itself and justifies its inclusion in the VR interface;

    a functionality that is missing in CAD systems. Also, the

    more breaks designers take, either voluntarily or involun-

    tarily, within the design activity, the less productive they

    are, e.g. getting lost in menus, choosing wrong parts (Pair

    8: q = 0.59, p < 0.1). This is explained in terms ofdesigners being interrupted by unproductive activity and

    then having to revaluate what they are doing before

    continuing with the design process (Mentis 2004).

    Pair 9 demonstrates that the designers were actually

    referring to the task information instructions rather than

    general help about the system when in the menu (Pair 9:

    q = 0.70, p < 0.05). This supports interview feedbackfrom the designers in which they said that the VR system

    was intuitive and easy to learn because they were focussing

    Table 4 Non-parametriccorrelations for activity and

    environmental categorisations

    Pair Variable (%) By variable (%) Spearman q Prob > |Rho|

    1 Model Design support 0.8909 0.0005

    2 Menu Design support 0.6848 0.0289

    3 Help/information Design support 0.8545 0.0016

    4 Menu Model 0.8788 0.0008

    5 Help/information Model 0.9030 0.0003

    6 Unproductive activity Menu 0.6848 0.0289

    7 Unproductive activity Drag and drop (cable point edit) 0.6121 0.0600

    8 Unproductive activity Sequence breaks 0.5879 0.0739

    9 Information Menu 0.6970 0.0251

    10 Drag and drop (cable point edit) Model 0.6364 0.0479

    11 Drag and drop (cable point edit) Menu 0.8303 0.0029

    12 Information Model 0.9273 0.0001

    13 System operation Designgoal 0.7939 0.0061

    14 Sequence breaks Drag and drop (cable point edit) 0.6727 0.0330

    Virtual Reality (2007) 11:261273 269


  • on the task in hand. This is further supported by the fact

    that when the engineers were in the model they were not

    looking at the task instructions and vice versa (Pair 12:

    q = 0.93, p < 0.05). Pairs 10 (Pair 10: q = 0.64,p < 0.05) and 11 (Pair 11: q = 0.83, p < 0.05) show thatmore time in the model usually means more time carrying

    out drag-and-drop activities (i.e. amending the design) and

    more time in the menus means less drag-and-drop time,

    respectively. However, more time operating the system,

    e.g. menu navigation, filter selections, etc., negatively

    impacted on creative design time (Pair 13: q = 0.79,p < 0.05) which suggests the need for well designed and

    specialised task-specific interfaces for all system opera-

    tions within immersive VR HMD design applications. This

    is also the case when more time spent in inter-sequence

    breaks meant less creative design time; in particular drag-

    and-drop (Pair 14: q = 0.67, p < 0.05).

    5 Cable harness assembly and cable harness

    installation planning

    One of the major benefits of any CAD system is the gen-

    eration of downstream manufacturing information; how-

    ever, in the area of assembly planning there is a

    considerable need for direct user assembly instruction data

    input. The nature of a typical CAD interface is such that

    there are considerable interruptions to the assembly plan-

    ners creative thought processes as they are generating

    these sequences which could affect the quality of the

    assembly plan output. What is required is a more intuitive

    method of generating plans in which the user can describe

    their assembly activities intuitively through actions and

    demonstrations of processes rather than explicitly

    describing them. It is within this context that immersive

    VR has an important role to play. Therefore, subsequent to

    the analysis of the design data in this study it was decided

    to investigate the possibility of automatically generating

    useable assembly plans using the immersive VR interface

    by firstly allowing the manufacturing planner to intuitively

    explore the cable harness and associated connector geom-

    etries to demonstrate a cable harness assembly process

    planning and then to follow this up by indicating the

    installing of the cable harness in a virtual model. As the

    user is logged, this automatically generates an installation

    assembly plan from the data which requires no interactive

    creation of instructions or subsequent amendments from

    the user.

    This approach supports the work carried out by Ritchie

    et al. (1999) which demonstrated that it was possible to

    produce production assembly plans via an immersive VR

    interface. However, work by Dewar et al. (1997) showed

    that the time achieved for virtual assembly planning was

    quite different from those obtained in the real world. This

    is a major disadvantage in virtual assembly planning

    environments. Therefore, to overcome this, the system was

    extended to demonstrate how real world assembly times

    can be cross-referred onto virtual equivalent tasks in a

    virtual planning environment. Tables of standard assembly

    times for fitting connectors to harnesses and fitting harness

    connectors into bulkhead connectors were tabulated from

    real world method studies and applied to the equivalent

    virtual tasks as an expert assembly planner built the virtual

    product. Non-intrusive logging of the planner enabled the

    development and generation of production-readable

    assembly plans without the need for human intervention, a

    major benefit over CAD methods. As well as this, harness

    access could be checked ergonomically.

    Once the domain expert was immersed in the virtual

    environment they were able to navigate around the cable

    harness and ergonomically and chronologically choose

    which connectors and cables to join together, thus facili-

    tating actual harness build. As this was being carried out,

    the user was non-intrusively logged in the normal way,

    connectors and cables identified and real world times

    automatically allocated to the sequence of build detected

    by the system. A similar approach was used when installing

    the cable harness into the actual assembly itself. The

    interface for assembly planning is shown in Fig. 11, and

    the assembly plans automatically generated for both the

    Fig. 11 VR user interface for assembly planning

    270 Virtual Reality (2007) 11:261273


  • harness itself and its installation, along with the corre-

    sponding real-world assembly times for each operation, are

    shown in Fig. 12.

    These outputs show that real world plans can be gen-

    erated automatically from user interaction within immer-

    sive VR design and planning systems. However, the

    matching of real world times with the virtual-equivalent

    activities demonstrates wider and more profound concepts.

    For example, interactive systems of this kind could be used

    in generic project planning domains by carrying out

    interactive assembly/disassembly in exactly this way

    (Gardiner and Ritchie 1999) and demonstrate the potential

    for generating data which could form the basis for for-

    malizing manufacturing intent.


    CABLE HARNESS BUILDING SEQUENCE--------------------------------------------------------

    14.24Hand Assemblyand inline connector CON21 (Type: socket Shell size: 1 Number of poles 2)Cable Bench40

    15.70Hand Assembly

    Connect cable CAB01(Type: SINGLECORE Number of Cores: 1 Core Cross-Section: 4.8 Colour (RGB): 255,0,0) to inline connector CON22 (Type: plug Shell size: 1 Number of poles 2)

    Cable Bench30

    27.18Hand Assemblyand inline connector CON24 (Type: socket Shell size: 2 Number of poles 7)Cable Bench20

    10.3Hand Assembly

    Connect cable CAB02(Type: CONTROLCY Number of Cores: 7 Core Cross-Section: 1 Colour (RGB): 225,125,0) to inline connector CON23 (Type: plug Shell size: 2 Number of poles 7)

    Cable Bench10

    Assembly Time

    (s)ToolingAssembly InstructionsW/CentreOp



    INSTALL CABLE HARNESS ASSEMBLY INTO EQUIPMENT--------------------------------------------------------------------------------

    7.69Hand Assembly

    Connect inline connector CON24 (Type: socket Shell size: 2 Number of poles 7) to bulkhead connector CON10 (Type: plug Shell size: 2 Number of poles 7) located at position (-2250,-500,-2175) and Orientation (-0,-1,-0,4.37114e-08)

    Assy Station40

    4.58Hand Assembly

    Connect inline connector CON23 (Type: plug Shell size: 2 Number of poles 7) to bulkhead connector CON05 (Type: socket Shell size: 2 Number of poles 7) located at position (1750,-500,325) and Orientation (-0,-1,-0,4.37114e-08)

    Assy Station30

    6.75Hand Assembly

    Connect inline connector CON22 (Type: plug Shell size: 1 Number of poles 2) to bulkhead Connector CON04 (Type: socket Shell size: 1 Number of poles 2) located at position (2250,-500,325) and Orientation (-0,-1,-0,4.37114e-08)

    Assy Station20

    7.17Hand Assembly

    Connect inline connector CON21 (Type: socket Shell size: 1 Number of poles2) to bulkhead connector CON01 (Type: plug Shell size: 1 Number of poles 2) located at position (3250,- 500,3725) and Orientation (0,-0,0.707107,0.707107)

    Assy Station10

    Assembly Time

    (s)ToolingAssembly InstructionW/CentreOp



    STANDARD REAL WORLD ASSEMBLY TIMES FOR EACH COMPONENT--------------------------------------------------------------------------------------------------

    Component Assembly Time (s) Component Assembly Time (s)CAB02 10.30 CON22 22.44CON23 14.88 CON21 21.42CON24 34.87 CON01 7.170CAB01 15.70 CON04 6.74CON24 34.87 CON05 4.58

    Fig. 12 Assembly plans for building cable harness and installing cable harness generated from assembly planner logging in the virtualenvironment

    Virtual Reality (2007) 11:261273 271


  • 6 Conclusions

    The novel outputs from this research have shown that it is

    possible to design and plan the assembly and installation

    of cable harness assemblies in immersive VEs using

    HMDs. It is also possible to examine, categorise and

    measure the wide range of design activities carried out by

    cable harness design engineers; something which has not

    been done to this level in the past. These novel categor-

    isations, along with their subsequent analyses, have pro-

    vided a more detailed understanding of design methods in

    this domain and a detailed outline of which aspects of VR

    are being used and where to focus future system devel-

    opment effort to improve performance. A numerical and

    statistical breakdown of activities has also shown to be

    possible which has given an insight into the cause and

    effect relationships taking place within the cable harness

    design process itself. The bona fide nature of these

    comparisons, expressing some of the oft-stated folklore

    relating to the design process, establishes that it is pos-

    sible to quantify the extent of the relationship between

    two or more subtasks. In the context of cable harness

    design, this analysis indicates that the categories chosen

    are valid and relevant to the general design function and

    could lead to a formalised standard and methodology for

    the analysis of creative computer-based design processes

    in the future.

    As a consequence of these findings, this research is

    being extended to apply the categorisation scheme within

    cable harness CAD design environments for direct com-

    parison with VR functionality as well as being used for the

    acquisition and formalisation of design ontologies related

    to cable harness design strategies and solutions. One area

    where this approach might prove useful is that thinking

    time could be extracted from the data, which in turn may

    spawn a capability to imply design intent from actions

    leading up to and after a decision making event. For cable

    harness routing, VR can give productivity gains over CAD

    (Holt et al. 2004); however, a more detailed investigation

    of cable harness design activities will be necessary to

    determine which tasks are best suited to VR and which are

    best suited to CAD. The design categorisation developed

    and successfully tested as part of this research is central to

    such an investigation.

    Areas for interface improvement have been identified

    from these experiments in terms of improving navigation

    and menu design, although they are not reported in detail in

    this paper. Once improvements have been implemented,

    the effects of these changes to the systems usability and

    functionality can be measured against the benchmarks

    reported here by reusing the categorisation scheme.

    Finally, assembly planning sequences along with the

    novel application of associated real-world assembly times

    can be generated by non-intrusively monitoring and log-

    ging the user. From this study, improvements in assembly

    planning interface design are being planned to make the

    assembly and disassembly of cable harnesses more realis-


    Acknowledgments We would like to acknowledge funding of thiswork by the UK Engineering and Physical Sciences Research Council

    Innovative Manufacturing Research Centre at Heriot-Watt University

    (The Scottish Manufacturing Institute) as well as the numerous

    industrial partners involved in the project.


    Adelson B, Solway E (1985) The role of domain experience in

    software design. IEEE Trans Softw Eng 11:13511360

    Akin O, Lin C (1995) Design protocol data and novel design

    decisions. Design Stud 16(2):211236

    Caudell TP, Mizell DW (1992) Augmented reality: an application of

    heads-up display technology to manual manufacturing processes.

    In: IEEE Hawaii international conference on system sciences,

    vol 2, pp 659669

    ChiCheng C, Jianzhong M, Gadh R (2002) A quantitative analysis on

    virtual reality-based computer aided design system interfaces.

    Comput Inf Sci Eng 2:216223

    Conru AB (1993) Computational support for interactive cable harness

    routing and design. In: 19th Annual ASME design automation

    conference, pp 551558

    Conru AB (1994) A genetic approach to the cable harness routing

    problem. In: IEEE conference on evolutionary computation, pp


    Cross N (2001) Design cognition: results from protocol and other

    empirical studies of design activity. In: Eastman C, McCracken

    M, Newstetter W (eds) Design knowing and learning: cognition

    in design education. Elsevier, Amsterdam, pp 79103

    Cruz-Neira C, Sandin DJ, DeFanti TA, Kenyon RV, Hart JC (1992)

    The CAVE: audio visual experience automatic virtual environ-

    ment. Commun ACM 35(6):6572

    Dani TH, Gadh R (1997) Creation of concept shape design via

    a virtual reality interface. Comput Aided Design 29(8):555


    Dewar RG, Ritchie JM, Carpenter ID, Simmons JEL (1997) Tools for

    assembly in a virtual environment. In: Proceedings of ICMA 97,

    University of Hong Kong, pp 583587

    Fischer AG, Chipperfield KC, Vance JM (2002) VR Hose: hydraulic

    hose routing in virtual reality with JackTM. In: AIAA/ISSMO

    symposium on multidisciplinary analysis and optimization,


    Gardiner PD, Ritchie JM (1999) Project planning in a virtual world:

    information management metamorphosis or technology going

    too far. Inf Manage 19(6):485494

    Gerber D, Bechmann D (2004) Design and valuation of the ring menu

    in virtual environments. In: Immersive projection technologies

    IPT 2004, Ames

    Gomes de Sa G, Zachmann G (1999) Virtual Reality as a tool for

    verification of assembly and maintenance processes. Comput

    Graph J 23:389403

    Guindon R (1990) Designing the design process: exploiting oppor-

    tunistic thoughts. Hum Comput Interact 5(2):305344

    Guindon R, Krasner H, Curtis B (1987) Breakdowns and processes

    during the early activities of software design by professionals.

    In: Empirical studies of programmerssecond workshop

    (December) Washington, DC, pp 6582

    272 Virtual Reality (2007) 11:261273


  • Holt POB, Russell GT (1999) Psychology as a science of design in

    engineering. In: Proceedings of the British Psychological Society


    Holt POB, Ritchie JM, Day PN et al (2004) Immersive virtual reality

    in manufacturing: design metaphors and cognitive ergonomics.

    Comput Inf Sci Eng 4(3):161-170

    Howarth PA (1997) The multi-factorial causes of virtual simulation

    sickness. Loughborough University, VISERG

    Howarth PA (1999) Oculomotor changes within virtual environments.

    Appl Ergon 30:5967

    Howarth PA, Costello PJ (1996) The nauseogenicity of using a head-

    mounted display, configured as a personal viewing system, for an

    hour. In: FIVE 96, Pisa, Italy

    Howarth PA, Costello PJ (1997) The occurrence of virtual simulation

    sickness symptoms when an HMD was used as a personal

    viewing system. Displays 18:107116

    Jayaram S, Vance J, Gadh R, Jayaram U, Srinivasan H (2001)

    Assessment of VR technology and its application to engineering

    problems. Comput Inf Sci Eng 1:7283

    Liang J, Green M (1994) JDCAD: a highly interactive 3D modeling

    system. Comput Graph ACM 18(4):499506

    McPhee K (1997) Design theory and software design, technical report

    TR 96-26. Department of Computer Science, University of

    Alberta, Canada

    Mentis HM (2004) Occurrence of frustration in humancomputer

    interaction: the affect of interrupting cognitive flow. Master

    Thesis, Cornell University, Ithaca

    Ng FM (1999) Virtual reality and computer-based tools for the

    routing of cable harnesses. PhD Thesis, Heriot-Watt University,


    Ng FM, Ritchie JM, Simmons JEL, Dewar RG (2000) Designing

    cable harness assemblies in virtual environments. Mater Process

    Technol 107:3743

    Preece J, Rogers Y, Sharp H, Benyon D, Holland S, Carey T (1994)

    Humancomputer interaction. Addison-Wesley, Wokingham

    Ritchie JM, Dewar RD, Simmons JEL (1999) The generation and

    practical use of plans for manual assembly using immersive

    virtual reality, engineering manufacture. Inst Mech Eng B


    Schon DA (1991) The reflective practitioner: how professionals think

    in action, Arena

    UK Health and Safety Executive (2000) HSE contract research report


    Varga E, Horvath I, Rusak Z, de Smit B, Broek H (2004) Survey and

    investigation of hand motion processing technologies for com-

    pliance with shape conceptualization. In: Proceedings of

    DETC04 ASME design engineering technical conferences and

    computers and information in engineering conference. ASME,

    DETC2004/CIE-57710, pp 114

    Weyrich M, Drews P (1999) An interactive environment for virtual

    manufacturing: the virtual workbench. Comput Ind 38:515

    Wolter J, Kroll E (1996) Towards assembly sequence planning with

    flexible parts. In: IEEE international conference on robotics and

    automation, IEEE, pp 15171524

    Zhu D, Latombe JC. (1991) Pipe Routing = Path Planning (with

    many constraints). In: IEEE international conference on robotics

    and automation, pp 19401947

    Virtual Reality (2007) 11:261273 273


    Cable harness design, assembly and installation planning using immersive virtual realityAbstractIntroductionImmersive virtual realityTypical virtual engineering applicationsResearch work domain

    Apparatus and methodologyApparatus: COSTAR experimental platform

    Experimental procedureAnalysis of resultsCable harness assembly and cable harness installation planningConclusionsAcknowledgmentsReferences

    /ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 150 /GrayImageDepth -1 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 600 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputCondition () /PDFXRegistryName ( /PDFXTrapped /False

    /Description >>> setdistillerparams> setpagedevice


View more >