The system should provide different ways to trigger processing these can be:

• Manual data workflow selection through the web interface
• Starting processing from other program/script through an API
• Automatic triggers:
  • New dataset appears in earlier specified storage locations