Long Term Knowledge Retention and Preservation Aziz Bouras University of Lyon, DISP Laboratory France abdelaziz.bouras@univ-lyon2.fr
Recent years: How should digital 3D data and multimedia information be represented so it will be readable X years from now? Conceptual Design (Knowledge-based CAD) Traditional CAD Analysis Virtual Prototype Environment QuickTime and a Cinepak decompressor are needed to see this picture. Engineering Specificatio ns Case Studies Component Data Other Resource Data. Design Repository Production Planning Version Revision Geometry Rationale. Design Evolution Database Process Planning Client/User Immersive CAD
From collaborative CAD to the whole lifecycle! Huge number of information systems Commercial Logistics Production ERP Quality Maintenance - ASS C.R.M. S.R.M. M.E.S. A.P.S. CAD C.F.D. PDM/PLM Serious Games C.A.E. IN.A.O. K.B.E. Marketing Methods analysis Industrialisation Quality assurance
Data creation Product creation process Data usage Enterprise BOM Enterprise PDM CAD CAE CAP E/E/S CAD: E/E/S: CAE: CAP: BOM: Computer Aided Design Electrics/Electronics/Software Computer Aided Engineering Computer Aided Planning Bill of Material
Problems with level of formalisms exist Ad hoc Hierarchies (Yahoo!) Terms Thesauri Structured Glossaries XML DTDs XML Schem a Formal Taxonomies Description Logics (DAML+OIL ) Ordinary Glossaries Data Dictionaries (EDI) Principled, informal hierarchies DB Schema Data Models (UML, STEP) Frames (OKBC) Glossaries & Data Dictionaries Thesauri, Taxonomies MetaData, XML Schemas, & Data Models Formal Ontologies & Inference Ack: M. Grunninger
Knowledge Preservation Context Why worry about long term knowledge preservation? Product lifecycles can be much longer than lifecycles for computing hardware, applications, storage media Enterprise domain-applications are becoming massive data and information-based European Commission and governments are imposing very strict preservation regulations (90years for aerospace, 20years for automotive ) Challenges Knowledge identification and extraction from huge heterogeneous warehouses and multimedia documents Future extensibility and reusability of digital preservation models Knowledge lifecycle assessment Interoperability of digital preservation platforms
i) today s starting point 9/6/2011 7
functionalities for long term digital preservation are currently being proposed Interaction between research clusters (GOSPI, NIST MEL) with industrial associations and big companies in US and EU: NIST(Nasa, Boeing, Ford, Nara, Darpa), MICADO(Airbus, Renault, Volvo), ENE (SMEs) 9/6/2011 8
High level recommendations for Open Archival Information System (OAIS) framework exist P R O D U C E R SIP Ingest Preservation Planning Data Management Descriptive Info. Archival Storage AIP Access queries result sets orders DIP C O N S U M E R Administration SIP = Submission Information Package AIP = Archival Information Package DIP = Dissemination Information Package MANAGEMENT
Gaps exist between long term preservation requirements and existing technologies KM approaches KM tools Task 3 Task 2 Task 1 Digital preservation Tasks, processes LTKR Requirements out of existing digital preservation tools: - Standards for archival system - Knowledge transmutation (dynamic features) - Agility in scope of long term Documentation and models (products history, web annotations, browsing information ) Systems Organisation (structure, history, reputation ) Ex. of current initiatives: LOTAR (STEP) for 3D engineering data preservation 9/6/2011 10
ii) future trends 9/6/2011 11
Build multi-layer architectures matching preservation functionalities Mediation layer: system connection and knowledge transfer Knowledge Integration System Connection Data Transfer Product, process organization knowledge Knowledge in Information Package form Knowledge Management Approach SIP DIP Data Preservation Planning Data Data Data Ingest Des.Info. AIP Data Management Archival Storage Administration Des.Info. AIP Access Information systems in enterprise used in product lifecycle Digital preservation platform: an extended OAIS 9/6/2011 12
Build mechanisms for knowledge evolution Build specific approaches and tools to regularly update the preserved knowledge (annotations, abstraction ) Test OAIS mechanisms and concepts (high level recommandations) Capture current corporate organization structure and strategy (formal information and internet) Context: knowledge identification ( Knowledge in enterprise, PPO model concept) Concept: knowledge integration for preservation (Knowledge model, OAIS concept) Design: architecture of KM (business view), service of KM (functional view) Implementation: digital preservation platform and connections (application/technical view) 9/6/2011 13
i) Priorities for the call 9/6/2011 14
From digital preservation perspective Analyse existing methodologies, standards and platforms for Knowledge Preservation as well as their capabilities to support the evolution of preserved knowledge Develop new programming models and scripts for knowledge extraction that hide the complexity of the product lifecycle underlying systems Provide new flexible approaches that allow users to extend existing functionalities to meet a variety of preservation requirements Analyse new massive data storage solutions such as cloud-scale services Analyse the impact on the current knowledge management approaches and possible re-engineering of current PLM/SCM systems (how the ontologies, metadata standards and registries developed today will be able to navigate the data warehouses of tomorrow) 9/6/2011 15
Form an enterprise perspective Identify generic requirements in terms of knowledge retention and archiving and propose application scenarios (through business processes identification) Define a pattern prototype for a knowledge preservation platform based on both structured information (Enterprise Information System) and non-structured information (data warehouse repositories, heterogeneous internet multimedia documents) Develop distributed data storage and processing systems on large clusters of shared commodity servers
Thank you for your attention! Contact: abdelaziz.bouras@univ-lyon2.fr +33 678 940 861