L2 EGs

• Describes real world

• Distinguished into

  • Data: stored symbols

  • Messages: stored concepts

  • Depend on context

• unit of information

• symbol fixed to a medium

• function which reduces magnitude of Data on a medium

• in conjunction with expansion method

• Data is reduced, some parts are deleted by:

  • Selection

    • identifying (ir-)relevant events

    • limited effectiveness or lost relevant data

  • Compression

    • requires knowledge of use case

• Data is reduced without losing data by:

  • (En-)coding

    • maximizing average data per symbol

    • reducing redundancy

• large amounts of data / growing data rates:

  • Development

  • Homologation (Zertifizierung)

  • Driving Function

  • Monitoring

  • Crash reconstruction

• e.g. Approval Trap overcome: 70 ExaByte necessary

  
  

precise knowledge is needed to differenciate between relevant and irrelevant data for selection

• Compression on selected information

• Fine tunable threshold for (ir-)relevance

• Lossless storage of relevant information

• lossless:

  • Huffman Coding: bit lengths for characters are determined by frequency of character

  • Dictionary coding: adding bits for recurring character combinations

• lossy:

  • VAAFO: human drives vehicle, AI acts virtually, identification of possible errors