Re: Virtual introductions re ODRL usage in the automotive world...

Thank you Benedict.

I am adding our mailing list instead of trying to accurately paraphrase
in a summary report and will do the same with Renato's message. We can
loop in the ODRL CG whenever you feel appropriate. I think we need to
come up with a clearer understanding of our needs first and will send
to the CG list.

Licensing and regulation are certainly factors for automotive. Data
collection will be using in-vehicle computing resources and bandwidth
to send to potentially multiple data warehouses. Those who incur those
costs would want to recoup them. Data providers will in many cases seek
compensation from the data consumers they are providing information.
This information is deemed personal and highly regulated especially in
the EU with GDPR. Even if there is no payment required for accessing
information, the data consumer should be bound by license what they can
redistribute to other parties. 

Currently the telematics service providers are going the secret sauce
route and that makes sense at present. Going forward will be different
with the common data model our spec references, the OBD2 data port
possibly going away and regulators likely requiring some information as
the do for heavy vehicles not to mention all the business interests in
this information.

I reached out to Eric Prud'hommeaux to learn what they might be using
for in the Healthcare and Life Sciences area. Challenge he conveyed in
using policy languages for capturing user (patient) consent is
communicating clearly and concisely the nuances of intricate
permissions pertaining to specific data points and their permitted
usage. We have the same need in automotive when dealing with an end
user (driver). Fleet management companies will be able to handle
reviewing more complex agreements.

On Tue, 2018-07-17 at 11:07 +0000,
benedict.whittamsmith@thomsonreuters.com wrote:
> Hi Ted,
> 
> Thanks for getting in touch. I'm really excited about the challenges
> the automotive sector will bring to big data.
> 
> The 'compare XACML and ODRL' question comes up a lot. So here's my
> take:
> 
> XACML models access control decisions. ODRL models licenses and
> regulations.
>  
> Imagine you had a complex, UNIX-style access control system and you
> wanted to make some sense of it: use XACML.
>  
> Imagine you had a bunch on licenses or regulations you wanted to
> express and enforce: use ODRL.
>  
> Our business problem at TR is to map from licenses and regulations to
> access control decisions. To understand the licenses and regulations
> we need ODRL. XACML has nothing to say.
>  
> To then make the access control decisions we could use XACML, but
> ODRL has already taken us most of the way. The XACML is now repeating
> the parts of the ODRL relevant to access control decisions, just in a
> different/parallel formalism.
>  
> But the parts of XACML that ODRL does not touch are the architectural
> elements: PDPs, PEPs, PIPs, … Here ODRL has nothing to say, so we're
> borrowing from XACML.
> 
> Both standards need to be situated within wider information models.
> We have one that formalises the relationship between:
> * data resources;
> * the supplier agreements and regulations that control their use;
> * the content services we provide our customers;
> * and the products we sell to them.
> 
> This allows us to guarantee compliance.
> 
> We have another that maps the permissions described by ODRL and the
> actual underlying data models that physically describe the data
> (stream topics, database rows ...).
> 
> This allows us to automate access control.
> 
> We're just at the start of this. TR is teetering on the edge of
> looking to standardise all these models across the financial data
> supply chain. But it's the age old debate: secret sauce or wider
> enabler.
> 
> Hope this helps. Happy to answer any further questions,
> 
> Ben
> ________________________________________
> From: Ted Guild [ted@w3.org]
> Sent: 16 July 2018 18:33
> To: Renato Iannella
> Cc: Whittam Smith, Benedict (TR Technology & Ops); Ivan Herman
> Subject: Re: Virtual introductions re ODRL usage in the automotive
> world...
> 
> I am also reading up on XACML to try to determine what might be more
> suitable to our needs. I would be interested to learn your
> impressions
> on how they compare.
> 
> Also let me know if you are alright with me adding a public mailing
> list to the cc.
> 
> On Wed, 2018-07-11 at 09:12 -0400, Ted Guild wrote:
> > Thank you Ivan.
> > 
> > Hello Renato and Benedict,
> > 
> > Forgive the delay on the thread, timing was before a few days off
> > and
> > it got buried.
> > 
> > This should be a relatively quick read to give you some
> > background on a task force we recently started. There have only
> > been a few calls so far. We already have some pieces of the big
> > data puzzle for automotive.
> > 
> > https://www.w3.org/community/autowebplatform/2018/06/18/w3c-automot
> > iv
> > e-
> > big-data-task-force/
> > 
> > Wiki has the areas under consideration for exploration listed.
> > Presently most of the interest is around consent capture, data
> > contracts and ontology work on the data model.
> > 
> > https://www.w3.org/community/autowebplatform/wiki/Data_tf
> > 
> > I was encouraged to take a look at ODRL and met with Ivan. It
> > looks very pertinent for granularly representing the types of
> > relationships we want for indicating what information can be
> > shared with which third parties plus some of our other potential
> > needs. My understanding from Ivan is customized profiles for
> > different purposes are necessary and you might be able to point
> > me to one.
> > 
> > The impression Ivan had is that ODRL has seen uptake in news and
> > media industry. Is there wider interest or any more generic and
> > ideally open source implementations? Commercial options also of
> > interest but experimenting would be facilitated by something more
> > readily available.
> > 
> > I am pleased to hear the work on a next version is continuing in a
> > CG.
> > If this seems pertinent, or to explore further if it might be, some
> > of
> > us would like to attend a CG call or invite you to one of our task
> > force calls.
> > 
> > Any examples and insight appreciated.
> > 
> > Regards,
> > 
> > On Sat, 2018-06-30 at 21:20 +1000, Renato Iannella wrote:
> > > Hi Ted, we would be happy to help out…
> > > 
> > > As Ivan said, we are “back” now in the ODRL Community Group and
> > > revamping efforts to promote ODRL V2.2
> > > 
> > > Please let us know (email to the CG) what areas your group would
> > > like
> > > assistance/feedback on..
> > > 
> > > Cheers - Renato
> > > 
> > > > On 29 Jun 2018, at 23:36, Ivan Herman <ivan@w3.org> wrote:
> > > > 
> > > > Renato, Ben,
> > > > 
> > > > please meet my colleague at W3C, Ted Guild, who is leading the
> > > > work
> > > > cooperating with the Automotive Industry at W3C. They recently
> > > > had
> > > > some work starting up that may lead to issues around data
> > > > access
> > > > and usage rights, which led them to a possible usage of ODRL. I
> > > > gave him some information, but I think you two are way better
> > > > positioned as I am as for the way it could be used in such a
> > > > different area (I told Ted about the profiling mechanism, but I
> > > > did
> > > > not find any reference to an existing profile like IPTC's…). He
> > > > was
> > > > also wondering whether there are general implementations around
> > > > that their community can use.
> > > > 
> > > > I let him ask the more specific questions…
> > > > 
> > > > Ted,
> > > > 
> > > > Renato and Ben were the Working Group chairs of the PO&E
> > > > Working
> > > > Group that standardized ODRL; ODRL is now 'back' in the hands
> > > > of
> > > > the ODRL Community Group[3] which is also chaired by Renato.
> > > > They
> > > > are in a much better position than I am in providing you with
> > > > up-
> > > > to-date information…
> > > > 
> > > > Cheers
> > > > 
> > > > Ivan
> > > > 
> > > > 
> > > > 
> > > > [1] https://www.w3.org/community/autowebplatform/2018/06/18/w3c
> > > > -a
> > > > ut
> > > > omotive-big-data-task-force/
> > > > [2] https://www.w3.org/community/autowebplatform/wiki/Data_tf
> > > > [3] https://www.w3.org/community/odrl/
> > > > 
> > > > ----
> > > > Ivan Herman, W3C
> > > > Publishing@W3C Technical Lead
> > > > Home: http://www.w3.org/People/Ivan/
> > > > mobile: +31-641044153
> > > > ORCID ID: https://urldefense.proofpoint.com/v2/url?u=https-3A__
> > > > orcid.org_0000-2D0003-2D0782-2D2704&d=DwIDaQ&c=4ZIZThykDLcoWk-
> > > > GVjSLmy8-
> > > > 1Cr1I4FWIvbLFebwKgY&r=6zKsY0rYbamT39VLwFyiNRQq5kP9V8VfXNbeD0Gwl
> > > > qvjGIQo1uv383jKtxD77PlM&m=Iiiwhjr938Qw9uP0owyLPzm5oy6HmdBASZvpZ
> > > > CaE0Rw&s=j9cnH1h-mnEhISQ1reeNUJ9IX7IKnpJFeKRL_GvcT0g&e=
> > > > 
> > > 
> > > 
> 
> --
> Ted Guild <ted@w3.org>
> W3C Automotive Lead
> http://www.w3.org
> 
> 
-- 
Ted Guild <ted@w3.org>
W3C Automotive Lead
http://www.w3.org

Received on Tuesday, 17 July 2018 15:47:18 UTC