DM: Piece can be part of multiple shipments, change from n:1 to n:n

[nscheiber-champ]

When mapping FWB and FHL to 1R, the following issue may arise:

• MAWB may have a different amount pieces than the sum of all HAWBs
• HAWB has the "real" number of pieces
• However, pieces of MAWB are in contractual lead for air transport

This leads to the following modelling problem:

• Both, MAWB and HAWB, need Shipment and separate Piece LOs (skeletons)
• Violates physics-centricity
• Violates single source of truth
• Business implication: On which Pieces to put LogisticsEvents? Or Checks? Or IoT devices?

In order to mitigate that issue, we need to be able to use the same Piece LO (or skeletons) for both MAWB and HAWB:

• Overlapping Piece LOs are connected to Shipment LOs of HAWB and MAWB
• Non-overlapping Piece LOs are connected to Shipment LOs either HAWB and MAWB

Hence, I propose to lift the cardinality restriction of Piece#ofShipment from n:1 to n:n.

[lambertciata]

Hi Niclas,
Indeed the Piece (n) <-> Shipment (1) relationship can cause issue if we have House declared.

With current situation we can only have the following:

While if we allow n:n relationship we can have the following:

This seems more accurate.