Persistently identify a dataset for human and machine use

RDA Group: 
Provenance Patterns WG
Nicholas Car
Data user
Repository manager
Data contributor
To be able to refer to a registered item in either a human- or machine-readable manner, share that reference and know that it will persist with useful functionality over perhaps a decade.
In order to be able to refer to an item with certainty that the reference will remain valid over time, I need to have it persistently identified. Additionally, I would like some basic metadata about it to be persisted too.
The item is identified with an object identifier
The object identifier can, in some system, be used as a key to look up metadata about the item
I can use a persistent identifier in a manner like an academic paper citation for human use
I can use a persistent identifier in a machine-redable way for structured metadata retrieval
Visit the landing page of the item
Instantly/easily obtain its persistent identifier
Instantly/easily obtain a citation format suggestion, incorporating that identifier, for human use
Alternative Steps: 
Visit the landing page of the item emulating a machine request
In a standardised way, obtain the persistent identifier for the item
In a stnadardised way, obtain basic, stnadardised, metadata about the item