Try   HackMD

Safran medium/long-term collaboration

Engineering 4.0

The MUSCLOR project focuses on modern ways to design complex systems (turbofans, landing gears, etc.).

Today:

  • a lot of tools, one/multiple per field of competence, totally disconnected
  • process: extract from a tool, send spreadsheets of data, import in another tool, version sometimes
  • more integrated conception = chaining tools
    • quick win approach, automate the existing process
    • a lot of work done to translate data (e.g. naming, units) from a tool to another

Target:

  • think new process/tools together
  • design future tools to be capable of better integration and allow better collaboration
    • used separately in a still sequential approach (inputs/outputs standards)
    • used together in a higher system level approach
      • assembly of the multi physics models to represent a part model that is not only a CAD
      • assess impact of one physic on the others (multi physics derivatives)
      • give access to this model to everyone, in read/write mode according their authorizations
    • no need to open multiple tools and do save/open several times
  • access shared information (CAD, physical data from simulations, etc.) but also modify it in a shared place
    • trigger actions to check validity of submitted modifications (clashed, max physical values on specific simulations, etc.)
    • trigger review from other teams (impact of your modifications on others)
  • adapt the view to the users: "Cockpit of conception"
    • a lot of different point of views are needed (engineering, technical management, etc.)
    • a single interface is maybe not the right approach
    • evaluate capability of a modular approach
      • through JupyterLab
      • e.g. an enginner may need the CAD editor module + domain specific simulation/viz tools (aero, mech, thermal, etc.), a marketing guy only a CAD viewer, etc.

Prototype this vision was the essence of CoSApp.

See also the great document of Benoît Combemale here.

tags: projects