RDF data handover and dereferenceable URIs

Hi All,

I am working on data handover architecture based on semantic web for EPC
(engineering procurement and construction) projects.

I have a question about handing over RDF data from one endpoint to other and
still maintain the dereferenceable URIs.

I want to know if other industries have come across such problems. If yes
then is there a “best practice” way to solve this problem?

For further details of the problem please consider following example
scenarios:

1.      There is an industry specific standard ontology. Building this
ontology is an ongoing activity. Each term goes from various stages of
standardization.

I.e. new terms (classes, properties) are submitted by various working groups
which host these terms at an endpoint under their control,

After initial screening these terms are moved to an industry consortia
endpoint,

After further review these are moved to an international standard endpoint.


All this time there are various application and projects using (referencing
to) these rdf entities.

Now the question is: how to keep a URI dereferenceable while it moves from
one physical endpoint to other through the standardization process?



2.      During construction of a large scale asset (like process plant or
oil refinery) there are various stakeholder companies involved. Main parties
are ‘owners-operator’ (someone who owns and operates the process plant) and
‘contractors’ (someone who builds the process plant for the owner). During
construction phase, contractor “owns” all the engineering data. With
implementation of semantic web technologies we envision he hosting this data
at an endpoint under his control. Since we need everything dereferenceable,
URIs used will be pointing to contractor hosted endpoint. At the end of the
construction phase, contractor hands over the facility (process plant) to
the owner. As part of the process, contractor also hands over the
engineering data to the owner. Owner then maintains the engineering data at
his endpoint for maintenance and operations activities.  In such case how to
keep the IDs resolvable when RDF data is moved from one endpoint to other?

I can think of some ways like maintaining a separate server to redirect
abstract URIs or using something like PURL (http://purl.org/docs/index.html)
but want to check with this group before committing  to a solution.

Thanks in advance for your help.

Regards,
rahul