lost in translation - society of american...

32
because good research needs good data Lost in Translation Technological Views on Preserving CAD Alex Ball DCC/UKOLN, University of Bath 15 August 2014 Except where otherwise stated, this work is licensed under the Creative Commons Attribution 4.0 International licence: http://creativecommons.org/licenses/by/4.0/ Supported by ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

Upload: others

Post on 14-Aug-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

  • because good research needs good data

    Lost in TranslationTechnological Views on Preserving CAD

    Alex Ball

    DCC/UKOLN, University of Bath

    15 August 2014

    Except where otherwise stated, this work is licensed underthe Creative Commons Attribution 4.0 International licence:http://creativecommons.org/licenses/by/4.0/

    Supported by

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

    http://creativecommons.org/licenses/by/4.0/

  • Outline

    Introduction

    3D geometric representations

    Advanced modelling techniques

    Use cases

    Possible solutions

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Introduction

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Key message

    The main technological barrier to preserving CAD isvarietyÉ of 3D geometric representationsÉ of advanced modelling techniquesÉ of use cases

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • 3D geometric representations

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Wire-frame modelling

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Wire-frame modelling

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Surface modelling

    Extruding Sweeping Lofting

    Revolving Triangular mesh NURBS

    cba

    Gre

    gA

    L

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

    http://en.wikipedia.org/wiki/File:NURBS_3-D_surface.gif

  • Constructive Solid Geometry

    + =

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Boundary representation

    =

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Mistranslation and misinterpretation

    c© Jianchang Qi, Vadim Shapiro

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

    http://dx.doi.org/10.1115/1.2218367

  • Modelling techniques: summary

    É There are many incompatible ways of modelling3D geometry.

    É There are many incompatible ways ofinterpreting 3D models.

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Advanced modelling techniques

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Construction history modelling

    1. Insert cylinder l = 20 r = 1.0

    Change cylinder l = 40 r = 0.5

    2. Insert sprocket r = 3.0

    3. Fit sprocket to cylinder

    4. Group cylinder and sprocket

    5. Scale group by 1.75×. . .

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Procedural modelling

    c© Yoav I. H. Parish & Pascal Müller

    c© Barbara M. Cutler

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

    http://dx.doi.org/10.1145/218380.218405http://people.csail.mit.edu/bmcutler/PROJECTS/PHD/

  • Parametric modelling

    8 cm

    28 cm

    8 cm

    18 cm

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Feature-based modelling

    RibsCavities

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Advanced modelling techniques

    É CAD models contain much more than justgeometry.

    É The geometry might be useless without theextra information.

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Use cases

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Reusing standard parts

    c© Koray Pekericli

    top-level assembly

    part f

    subassembly a

    part a part b part c

    subassembly b

    part d part e part f

    part j

    subassembly c

    part d part e part g

    subassembly d

    part a part h part i

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Relationship with other documents

    CADModel

    ArchaeologyDatabase

    Bill ofMaterials

    ProcessModel

    RationaleModel

    SystemsModels

    . . .

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Integration with other systems

    CADSystem

    CAMSystem

    FiniteElementAnalysisSystem

    GeographicInformation

    System

    AnimationSystem

    HighQuality

    Renderer

    . . .

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Use cases

    É You might need to coordinate CAD models withmany other types of information.

    É You might need to mimic whole systems.

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Possible solutions

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Standards: IGES

    2D drawing

    Wire-frames

    Surface modelling

    Constructivesolid geometry

    Boundaryrepresentation

    IGES(ANSI Y14.26M-1981, . . .

    ANS US/PRO/IPO-100-1996)

    WunderCAD CAD-

    ulous

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Standards: IGES

    2D drawing

    Wire-frames

    Surface modelling

    Constructivesolid geometry

    Boundaryrepresentation

    IGES(ANSI Y14.26M-1981, . . .

    ANS US/PRO/IPO-100-1996)

    WunderCAD CAD-

    ulous

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Standards: STEP

    STEPISO 10303?

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Standards: STEP

    STEPISO 10303!

    LOTAR(NAS 9300/EN 9300)

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Standards for BIM, CAD style conventions

    BIMÉ ISO 16739:2013 Industry Foundation Classes (IFC) for data

    sharing in the construction and facility management industriesÉ National BIM Standard – United States (NBIMS-US)É AEC (UK) BIM ProtocolÉ (BSI) PAS 1192-2 Information management for the

    capital/delivery phase of construction projects

    CAD styleÉ United States National CAD StandardÉ AEC (UK) CAD Standards For Layer Naming

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • Recommendations

    É Establish why a CAD model will be kept, then target therequired properties for preservation.

    É Create tests that can prove whether these properties havesurvived.

    É Keep native CAD models for as long as they can be read.É Normalise to STEP/IFC and a geometry-only standard (or two).É Don’t forget supporting documentation, especially local

    conventions and ‘house style’.É Campaign for better support for standard formats in CAD

    systems!

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

  • DPC Technology Watch Report

    01000100 01010000 01000011

    01000100

    01000100 01010000 01000011

    01000100 01010000 01000011

    PreservingComputer-AidedDesign (CAD)

    DPC Technology Watch Report 13-02 April 2013

    Alex Ball

    http://dx.doi.org/10.7207/twr13-02

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

    http://dx.doi.org/10.7207/twr13-02

  • because good research needs good data

    Thank you for your attention

    DCC Website: http://www.dcc.ac.uk/

    ‘Preserving CAD’ report:http://dx.doi.org/10.7207/twr13-02

    ARCHIVES*RECORDS 2014, Washington, DC 2014-08-15

    http://www.dcc.ac.uk/http://dx.doi.org/10.7207/twr13-02

    Introduction3D geometric representationsAdvanced modelling techniquesUse casesPossible solutions

    fd@510-Ball-Lost-in-Translation-1: fd@510-Ball-Lost-in-Translation-2: