Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: 13/4/2022 minutes
Quick links: [ CVII home page ] [ Tech Stack components ]

2022-04-13

Agenda:

  1. Project Status
  2. AMM Sessions


Minutes:

  • Project status roundup abbreviated to allow more AMM prep. Any
    • VISS
      • Ted: BMW hope to OSS
      by June (hope)
      • the VISS→Android adapter they will discuss at the VISS session at the AMM by June
      • Ulf is refactoring WAII so the architecture has a much lower number of standalone processes. Work should be complete in 2-3 weeks.
    • Kanto
      • Steve: Tom could you chase Tina for the Kanto presentation please.
    • Aos
      • Steve reports that the public documentation for Aos is now online https://docs.aoscloud.io
      • Work is starting on the previously reported goal of replacing the current VISS v1 server with an OSS replacement to adopt and extend. WAII and Kuksa.val in focus.
      • Work is also starting on some of the VSS related topics mentioned in Artem's presentation last year such as user service migration between cars.
  • AMM workshop Thursday afternoon: 2-3 hours
    • Discussion of what the group wants to work on and discuss in near-term
    • Currently the sessions are split into the two related topics VSS Data Storage, including feeders and VSS Data Servers
    • Topic brainstorming:
      • VSS Data Storage
    AMM 2-3 hours
    • VSS Data Storage
      • (including VSS Data Feeders) (Southbound)
        • Data Storage implementations:
          • DBs: redis, timeseries etc
        • Architecture: Single central vs distributed vs direct sensor connect (bypass store)
        • Feeders: CAN (kuksa.val
        advanced
        • have closed some gaps), some/ip,
        • Timeseries vs last/single value
        • Interfacing into data servers
      • VSS Data Servers (Northbound)
        • VISS v3?
        • Are VISS backend abstractions possible: storage, feeders?
        • GraphQL - in-vehicle?
        • VSS Data Servers to cloud
        • Data processing
          • Ted explains they see issues with OEM cloud data. Time slice not short enough for some forms of h/w maintenance wear tracking so have to do their own. Possible need for some in-vehicle processing. Also issues with in-vehicle caching with insufficent IO.
            • In-vehicle vs cloud
            • In-memory vs storage

2022-04-06

Agenda:

  1. Project Status
  2. AMM Sessions

...