IETF 118 - ASDF WG 2023-11-06

1. Note Well. https://www.ietf.org/about/note-well/

2. Logistics for Meeting

   1. Meetecho: https://meetecho.ietf.org/client/?session=31725
   2. Notes: https://codimd.ietf.org/notes-ietf-118-asdf
   3. Roll call/Blue: sheet covered by Meetecho + app

3. WG status update (Chairs - 5 min)

   Niklas presented the current situation.

4. SDF (Carsten - 20min)

   1. SDF overview
   2. Finalization and processing of last comments

   Carsten presented:

   Base SDF WGLC processing.

   Had WGLC, ended Sept 20th, got detailed comments, published -16, got
   some more comments, did -17 yesterday, some more editorial work
   needed, so -18 will be done next week. Intent to publish to IESG
   soon.

   Changes:

   • editorial + some new words (Groupings), some of the more useful
     terms already taken, shoul dbe consistent though.
   • new 3.4 about top level definitions
   • content format IANA assignment
   • added description in info block
   • JSON pointers or local references for sdfRequired (what other
     parts needed to build something, can now point to grouping or
     itself). Easier to manage.

   The Future:

   SDF Base is done. Looking at

   • SDF Mapping Files: augment base SDF spec to accomodate a
     particular ecosystem. Also allow translation (e.g., YANG to SDF
     translator)
   • SDF Compact Notation: Maybe informational
   • Links: Links are important, two levels:
   • links between sdf models (aka relationships)
   • sdfType for links
   • New (potential) stuff
     • NIPC
     • DT location extensions

   Carsten presented SDF Intro, same as at first ASDF meeting.
   https://datatracker.ietf.org/doc/slides-118-asdf-sdf-intro/

5. Potentially related drafts

   1. An Application Layer Interface for Non-IP device control (NIPC)
      (5 min)

   Bart Brinckman (Cisco) presented NIPC:
   NIPC aims to address with siloed IOT applications (own stack, own
   radios). A lot of infrastructure duplication. No clear NB interface.

   Problem: How do we onboard/control/receive telemetry? in this
   diverse scenario, in particular w non-IP devices.

   Aim to intrdocude standard APIs. Build connectivity metho/radio
   agnostic gateway with tech specific controllers.
   Three interfaces

   • onboarding - based on SCIM
   • device control - NIPC
   • telemetry - broadcasts, streaming data

   Implemented APIs from application perspective, in field test. Have
   open source applications (based on enterprise network + application
   side)

   Why SDF: If a model need to interoperate you need a way to connect,
   ie needs abstracted generic apis that works across multiple
   technologies.

   Michael: Are you proposing a standard serialization of SDF - Eliot:
   No. NIPC is on the other side of the bridge,

   Kevin Smith/Vodafone: Would it make sense to add NB-IOT. Potential
   alignment w non-IP standards work in ETSI

6. Extended information of Semantic Definition Format (SDF) for Digital
   Twin (5min)

   Hyunjeong Lee (ETRI) presents Extended info for SDF for DTs draft.

   SDF currently lacks location information. It is a key information
   for Digital Twins. Location needed to identify objects.

   proposal uses ISO 23247-3 as basis, wither for DT, personnel.
   Intent to extend SDF information with location, aka sdfLocation.

   David Navarro: why is this not a sdfObject?
   A: Location is not currently described in sdf

   Ari: DTs is a very good use case, it is about how to express it,
   have experiemented with SDF, potentially use Object.

   Summary: we want location info, not sure how to best represent it.

7. ASDF future and re-charter (25min)

   Charter discussion

   Michael: We have done the things in the charter. Things we have
   proposed earlier can be said to be covered already.

   Carsten: New proposals show that there are additional extensions
   into physical domain, maybe some things that can't be touched.
   Furthermore, we don't have a way on IP layer to work non-IP. Could
   be a basis for a program to do next sdf.

   Eliot: have a compelling use case for non-IP, means that we can
   extend the reach to additional devices. Propose to build the bridge.
   Look what exists before, and re-use.

   Niklas: Notes that location is closer to the current SDF abstraction
   level than NIPC, but both are relevant topics.

   Alexander Pelov: Wrt NIPC, problem to be solved there, lora, zigbee,
   etc. Need to start poking in that direction. Could the interface be
   described using SDF description.

   Ari: Would be useful to flesh out how concepts relate together: (1)
   NIPC is low level and (2) SDF is what data means.

   Michael: Reason to stay on semantic level was to stay out of other's
   domains. SDF would allow for mechanical translation, basically a
   developer level tool.

   Ari: Not sure if how it fits together, need to do whiteboarding.
   Won't jump to conclusions.

   David: SDF is defining a data model and a way to interact w the
   data, the interaction with ecosystems is through mappings.

   Eliot: would like to hold a side meeting to discuss this, there is a
   NIPC mailing list. Will email list with proposal for side meeting

   Michael proposes energy poll on active drafts:

   compact sdf: 5

   digital twin/location: 5

   NIPC: quite a few, more than can quickly count

   sdf relations: 2

   sdftype: 4

   Outcome: There is probably enough energy so that we might be able to
   get through the current documents in the list.

8. AoB

   nothing

Meeting closed