**Session Date/Time:** 16 Sep 2026 06:00

[00:00:06] **Ari Keränen**: Good morning.

[00:00:11] **Lorenzo Corneo**: Good morning or good afternoon.

[00:00:17] **Ari Keränen**: Good afternoon too.

[00:00:19] **Lorenzo Corneo**: Yeah. Yeah. Let me see if my slides are here. Yes. Yeah. Let let's wait a few more minute. We are expecting some more participants, and then we can start. Hello, Carsten and Rohit.

[00:01:52] **Rohit Mohan**: Hi.

[00:01:56] **Lorenzo Corneo**: Rohit, is Bart joining? Do you know?

[00:02:00] **Rohit Mohan**: I think he should be.

[00:02:02] **Lorenzo Corneo**: Okay. Then then we'll wait couple of more minutes for him to join, and then we can start. Thanks. Hello, Bart. Are you up and running, Bart?

[00:03:49] **Bart Brinckman**: Yes.

[00:03:50] **Lorenzo Corneo**: Awesome. Great. So I I think we can start. At least the presenters are all here. I'm not sure Michael joins today. He's pretty late there. So welcome everyone to this ASDF working group interim meeting. This is the sixth meeting of the year. So this is the preliminary agenda that we drafted together with Michael. And so the first time items would be kind of bringing up and addressing if there are any issues open in NIPC or in the SDF protocol mappings. And then we have the updates from the SDF Extension for Non-Affordance Information and the SDF modeling for Digital Twin (v05) drafts. Is there anything more that should be taken into account for the agenda today?

[00:05:02] **Hyunjeong Lee**: Can you hear me?

[00:05:03] **Lorenzo Corneo**: Yes.

[00:05:08] **Hyunjeong Lee**: Five.

[00:05:09] **Lorenzo Corneo**: Five. Sorry. Sorry. It was an old template. Apologies. Yes. Yes. My my bad. Okay. So I take it that, like, there there's nothing more in in the agenda. So maybe we can start. I guess it's gonna be Bart or Rohit, depending who wanna go first.

[00:05:37] **Rohit Mohan**: Yeah. I can just quickly update. Yeah. Yeah. So I think the draft numbers are wrong here, but I did publish a new version of SDF protocol mapping in NIPC today. I think it's 12 and Nipsey 22. The main updates are so I think there's a PR from Ari on, like, some text for handling SDF maps with protocol mapping. And another update that we did was we added a new event type for Zigbee in both drafts, which is like a it's a new IANA registration in the draft with, like, one of the extension points. I don't think we've received any comments for the IESG review, right, yet, right, like, from the AD. So I think that that's that's the only other update. I think one other update that we had is we've also added, like, an implementation section now to the SDF-four core mapping draft.

[00:06:41] **Lorenzo Corneo**: But yeah. Rohit, sorry to jump in. That is correct. There is not yet a review. Chidambaram is on vacation. And yeah. He said he he will do it as soon as he can. And and and, also, I apologize for the wrong numbering of the drafts. I just came back from vacation and so sorry. Yeah.

[00:07:03] **Rohit Mohan**: Think it's just an old old sled, I guess. Yeah.

[00:07:06] **Lorenzo Corneo**: It's it's yeah. At least I made it to the office. Okay. Sorry. Go ahead.

[00:07:12] **Rohit Mohan**: No. I think that that was it. I I don't think we had anything more.

[00:07:22] **Lorenzo Corneo**: Yeah. Yeah. I'll also we'll go through the the issues. We we have been discussing that Ari has been discussing that with me also. But, yeah, Ari, go ahead.

[00:07:37] **Ari Keränen**: Maybe a quick heads up on the protocol mapping since the original version that went for the AD review didn't contain the latest text, might want to now since the text is already in, you might want to remind the aid that, hey. By the way, we did this small one small update and give a, you know, quick rationale. It basically it's not non normative change, just giving you guidance how to use SDF in that specific case. So I think in that sense, it's a small change, but I think the AD would appreciate that heads up.

[00:08:07] **Rohit Mohan**: Okay. That's something that I have to sorry. I'm not familiar

[00:08:12] **Ari Keränen**: with Maybe that's from the chairs easiest. Yep. That's that's that's notifying. Hey. By the way, after pushing the button, we added this one small thing. I think chair chairs chairs is maybe the best since now it's in the 80 review already.

[00:08:34] **Lorenzo Corneo**: Yeah. Can do that. Yeah. Maybe Rohit send an email also to Michael to Ramon about that.

[00:08:44] **Rohit Mohan**: Yeah. I did send an email to Michael, like, before I published the draft, I think, to Michael.

[00:08:49] **Lorenzo Corneo**: Okay. Yeah. We got it. Okay. If that is the one, we we got it. Okay. Thanks. Yeah. Also, you reached out about the implementation. I I replied to you via email. So the thing with the implementation that we had, it was on, I think, NIPC version four. And I'm not sure the protocol mappings were already there. Okay. So yeah.

[00:09:21] **Rohit Mohan**: Okay. That's fine. We don't need to have that. Right now, we have the two Cisco implementations there. Actually, I I wanna check with Ari as well because I believe Ari had, like, the like, a SDF editor that supported the protocol map. So is that something that should go there? Yeah. I think, Ari. Yeah.

[00:09:41] **Ari Keränen**: Yeah. I haven't had a chance to update that to to use the protocol mapping yet. And and as as you noticed, now, the tool is is down because I haven't had a chance.

[00:09:49] **Rohit Mohan**: Yeah. I yeah. I couldn't check if it was there or not because it doesn't load.

[00:09:54] **Ari Keränen**: Yeah. Exactly. I mean, I the server needs a bit of maintenance. I typically use only for hackathons, and there hasn't been hackathon for a while, so I haven't had a chance to do the full maintenance there. I mean, of course, it would be easy to add to the to the editor because it all it requires is updating the JSON schema. But but maybe but I I I don't think that's really needed for for for this draft because I suppose your implementation already implements for the call map. So, I mean, there is an implementation of it. Yeah.

[00:10:31] **Rohit Mohan**: Yeah. We already have, I think, two implementations in the implementation stages section. Yeah. We're just wondering if there's any more that we need to mention. But, yeah, talk to you.

[00:10:41] **Ari Keränen**: Yeah. Thanks for checking. And I'm sorry I couldn't have a chance to reply to you earlier. But, yeah, I haven't implemented it, but I think it's that's okay. So no need to mention it.

[00:10:51] **Rohit Mohan**: Yeah.

[00:10:59] **Lorenzo Corneo**: Thanks, Rohit, for the for the update. Is there anything more?

[00:11:07] **Rohit Mohan**: I don't think so.

[00:11:10] **Lorenzo Corneo**: Okay. Yeah. Thanks for the update. Yep. Very happy the work is is going and it's proceeding also eventually in the queue. Thanks a lot for driving this, Rohit. I guess we can then move to NIPC updates with Bart.

[00:11:34] **Bart Brinckman**: Okay. So on the the NIPC side, I think Rohit already mentioned it. Actually, we we added this. We didn't make any changes other than this new event type we added for Zigbee. So the the goal there was we could always do this afterwards, but we thought, hey. This might be a good event type to add. And we thought it we we could put it in the text other than adding it later through Ayanna and Ayanna registration. So that's that's all we did on NIPC. And, again, we're waiting for Chidambaram's review.

[00:12:30] **Ari Keränen**: And I guess then also here, the same comment applies. You you wanna give a heads up that there were also here was a small change since the person that was posted for a review.

[00:12:39] **Bart Brinckman**: Yeah. The the, honestly, I mean, Rohit the email Rohit wrote included both. So if just forward the same thing to Chidambaram, then I think that he has all the information.

[00:12:52] **Lorenzo Corneo**: Yeah. Okay. We we we we'll take care of that myself and then

[00:13:01] **Bart Brinckman**: I can Mike. I'm happy to do it as well. It doesn't matter. But if you if you're if you're gonna do it, then that's fine.

[00:13:07] **Lorenzo Corneo**: Yeah. Yeah. I'm gonna do it. Yeah. Or maybe, actually, Michael has already done it. Still catching up with the emails, but we'll we'll take care of this. No need to bother.

[00:13:17] **Bart Brinckman**: Thank you.

[00:13:18] **Lorenzo Corneo**: Yeah. Sure. Thanks to you. It's great work. Do we have any anything more to discuss on on NIPC here? Otherwise, I'm happy to give the words to either Hyunjeong or Jungha. Yeah, Bart?

[00:13:37] **Bart Brinckman**: Yeah. I think we're yeah. Sorry. Yeah. No. I think we're good.

[00:13:39] **Lorenzo Corneo**: Okay. Yeah. Thanks for the great work, Bart and and Rohit. And yeah. So who wants to go first? Hyunjeong or Jungha?

[00:14:03] **Hyunjeong Lee**: I will present first. Okay.

[00:14:06] **Lorenzo Corneo**: Then I load your line. Did it something? Yes. There you go.

[00:14:20] **Hyunjeong Lee**: How? Try to share. Okay. Hello, everyone. Today, I present the recent updates to Sorry.

[00:14:31] **Lorenzo Corneo**: I have to pass you the control of the slides.

[00:14:35] **Hyunjeong Lee**: Okay. Okay. Thank you.

[00:14:38] **Lorenzo Corneo**: Yes.

[00:14:40] **Hyunjeong Lee**: Oh, hello, everyone. Today, I present the recent updates to SF modeling for digital twin version five. In this revision, we mainly clarified the protocol related descriptions, refined the domain examples, and added a new smart network use case. Next slide, please. I can okay. First, let me briefly summarize the major changes since draft form. In section five, we clarified the relationship between this document and the SDF protocol mapping specification. In particular, we now explicitly state that the current protocol mapping specification defines mappings for BLE and Zigbee. We also revised the section 5.4 to clarify the scope of a QoS. In section six, we reorganized the examples as the touchscreen systems in different application domains, and we added a new smart network system example as section 6.5. We also updated the terminologies to report to RFC ninety eight eighty, refresh re several references and made editorial improvements to the security and crucial sections. Table of contents. This this slide shows the structural changes between version four and five. The overall document structure remains mostly the same, but there are several important updates. First, section 5.4 was simplified from QoS and synchronization semantics to QoS considerations. Second, we added a new section 6.5 for the smart network system. We also corrected the section seven title and updated the references to reflect the latest SIF related specifications. So most of the changes are not complete restructuring of the document, but whether clarification and extension of the existing model. In this slide, one important change in draft version five is the clarification of protocol support. In the previous version, table two could be interpreted as if all the all the listed protocols were already defined by the protocol mapping specification in draft version five, we explicitly clarified that the current protocol mapping specification defines mappings for BLE and JigPE. At the same time, table two still includes other protocols such as MQTT, CoAP, HTTP, WebSocket, things like that because these possible protocols that may be used in data driven implementations. So the key point is that table two is a list of applicable protocol examples, not a list of protocols currently standardized by the SDF protocol mapping specification. We also refined the protocol binding examples. Figure three now shows BLE mapping using service ID and characteristic ID, which makes the basic example more consistent with the current SDF protocol mapping work. Another change is related to QoS. We renamed the section 5.4 to simply QoS considerations. More importantly, the QoS level zero, one, and two are now explicitly described as an MQTT exam. This avoids giving the imp impress impression that these QoS levels are common semantics for every protocol. In other words, 3.5 separates the general concept of QoS from the protocol specific QoS mechanisms or clearly. Section six

[00:19:46] **Lorenzo Corneo**: Hey, John. I I have a question. Sorry. I lost connection for for for a moment in in the previous slide.

[00:19:53] **Hyunjeong Lee**: Yeah.

[00:19:55] **Lorenzo Corneo**: So I I'm wondering, can you so the this quality of service feature, for example, could be found found in MQTT. Are you planning to insert that in SDF protocol mapping for MQTT, for example? Or is QoS outside the protocol mappings?

[00:20:21] **Hyunjeong Lee**: But I don't have any place to add MQTT, things like that. Just the list of protocols. The table shows the list of protocols.

[00:20:32] **Lorenzo Corneo**: Yes. Yes. But it is your idea to embed the quality of service inside MQTT protocol mapping, Or is it something because I'm I'm thinking if we can reuse the existing tech stack that we have in in in SDF.

[00:20:54] **Hyunjeong Lee**: I'm not sure, but the QoS consideration is one of the MQTT parameters. But, currently, I don't have any examples for the QoS considerations.

[00:21:09] **Lorenzo Corneo**: Okay. Okay. Because when when I read this slide, the first thing that I thought about, okay, but then we can have a protocol map for MQTT. And QoS would be one of the fields that you have in that object. Like, you have here BLE, you have service ID, you could have MQTT and then a field in the object called QoS, and then you can write zero one or two, for example. Just put your thought for for for for the drafts.

[00:21:43] **Hyunjeong Lee**: Oh, okay. Next time, I consider the NQTT QoS exam first.

[00:21:50] **Lorenzo Corneo**: Yes. Thanks. Also, I will go through the draft. And given that we have this, I want to remind everyone that now this draft is on our GitHub repository. So please go go in and then open issues or submit PRs, everybody.

[00:22:09] **Hyunjeong Lee**: Oh, okay. Yep. Okay. I thank you. I'll upload the latest version to the GitHub.

[00:22:18] **Lorenzo Corneo**: Sorry?

[00:22:19] **Hyunjeong Lee**: I'll upload the current version to the GitHub GitHub. Okay. Okay. Okay.

[00:22:29] **Lorenzo Corneo**: Thank you.

[00:22:31] **Hyunjeong Lee**: Thank you. But changes does not go up. I'd like to page I'd like to show the page six, but the page is not moved to page six.

[00:23:05] **Lorenzo Corneo**: Yeah. I I I see page six. Yes.

[00:23:09] **Hyunjeong Lee**: Oh, do you see the page six? Six. Oh, six. Okay. So in in section six, six section six has also been slightly re reframed. Instead of simply presenting a collection of examples, the section now presents data twin systems in various domains. The document currently includes four domains. The first is a modern system based on photo number seven. The second is a healthcare system based on the patient monitoring with the ECG and infusion pump components. And the third is the smart building system using lighting and occupancy information. And the fourth is the newly added smart network system. The purpose of these examples is to demonstrate that the same SDF based modeling approach can be applied consistently across different application domains. The most significant new example in draft-ietf-asdf-digital-twin is the smart network system in section 6.5. The basic flow is shown here. The first, the system obtains tracking information associated with a network device. Second, these are twins are created for both the network device and the traffic information. Third, the current traffic information from the network device is synchronized with the corresponding traffic information twin. Finally, the synchronized data twins can be used as input to train a prediction model for detecting or predicting on abnormal condition of the network device. In the SDF model, the network device can be represent as an SDF object. Tropic values can be represent using SDF property and relations between the twins can be expressed using a SDF relation. The model training itself is performed at the at the application layer while a SDF represents the synchronized data twin state used by the analytics function. In addition to these technical changes, we made several editorial and reference update. First, the base SDF specific Jungha? Creation.

[00:26:13] **Lorenzo Corneo**: Jungha, sorry to interrupt. We have a question from Ari, I guess, on the previous slide.

[00:26:19] **Ari Keränen**: Okay. Yes. Thank you. First of all, think this is great. I I like your approach of, like, applying the modeling technique across different domains and seeing that it scales. This is this is really, really useful. Can you talk a bit more about what is this smart network system object used? You mentioned about up abnormal condition detection. But but can you can you talk a bit more about the use case? Where is it actually used?

[00:26:47] **Hyunjeong Lee**: So abnormal condition is like a fire. In Korea, some years ago, there is a fire in the in the KT, Korean net Korea Telecom, the building. So I guess some scenarios for the fire or rain, things like that. Some kind of some kind of abnormal conditions. Is that okay to you?

[00:27:24] **Ari Keränen**: Okay. Yeah. I think I got can I get a gist of it? Okay. But I'll take a look. But thanks. This looks very interesting. And I'm glad to see that you you're able to map it well to kinda existing capabilities without needing additional. I think that that's a good sign that we start to have a a scalable set of of capabilities here. Thanks.

[00:27:44] **Hyunjeong Lee**: Thank you. Yeah.

[00:27:45] **Lorenzo Corneo**: I I I also have have a question here. So I see that you're using SDF relation, and I guess this is more a question for everybody in the audience. The draft is currently expired. It has not been been brought forward. What do we do about it? My question is for you, Hyunjeong, will your digital twin modeling still stand if Mhmm. SDF relation is not there, or is it necessary for you to have the SDF relation? I'm asking this because if it is necessary, then we have to do something about it. So resume the draft and then bring it to completion. That's my question.

[00:28:31] **Hyunjeong Lee**: Yes. So SDF relation is needed for my current document. So I and Jungha will okay. We'll update the SDF stipulation document if possible. We will consider that.

[00:28:48] **Lorenzo Corneo**: Okay. Okay. So you would be willing to do that?

[00:28:53] **Ari Keränen**: Yes. Awesome.

[00:28:55] **Lorenzo Corneo**: Thank you. Then then we will arrange that maybe offline. Yeah. That that's great to hear. Thanks. And then we have Ari. Ari, are you in the queue?

[00:29:15] **Ari Keränen**: Sorry. On the mute button. Yeah. Just let's just say yeah. Yeah. That's also, operate on the SDF-Rel is also something on on on my agenda. So I think we we should definitely revive it and and and and push it forward because it there seem to be quite a bit of use for it in different domains. This included. So, yeah, let let's do that.

[00:29:35] **Lorenzo Corneo**: Okay. Agreed. Then we also have some volunteers to to bring it forward. I'm I'm I'm very happy of that. Okay. Thank you, Hyunjeong. Please go ahead.

[00:29:46] **Hyunjeong Lee**: Okay. Thank you. Updates. In addition to these technical changes, we made several and reference update. Post the base is the specification is now referenced as RFC

[00:30:09] **Lorenzo Corneo**: And

[00:30:13] **Hyunjeong Lee**: those as a protocol making reference was also updated to the latest document. Latest trip to use in this revision, we refined the security considerations to more clearly mention of the authentication and authorization for dispatching context information. We also improved the wording in the implementation procedure, particularly around cons configuration, interworking, and synchronization. Overall, the draft-ietf-asdf-digital-twin is intended to read less like a simple collection of examples and more like a structured guide for a SDF based modeling. And next SDF. Finally, I'd like to brief briefly mention some possible next steps. One topic we are considering is how AI-based components could be interacted with the SDF based data twins. Currently, the document mainly focuses on how to represent and synchronize data twins. A possible future detection is to consider how synchronized data twin information can be used by an AI agent for instance for inference, decision making, and autonomous control. Conceptually, we can think of this as moving from a passive digital representation toward toward a more active setting. At this stage, this is only a possible direction for further discussion, and you would appreciate feedback from the working group. Thank you. Any comments or questions? And I have I have a question to Lorenzo. If the SF relation so do we need to contact the original editor for SF relation document?

[00:32:19] **Lorenzo Corneo**: Sure. You can do it, but he sits next to me. So I can also, you know, send send an email to to everyone involved, and and and we can figure it out. But, anyway, I already discussed with Ari, which is the original author, and he is very keen on having people taking over the work.

[00:32:46] **Hyunjeong Lee**: Okay. Thank you.

[00:32:47] **Lorenzo Corneo**: Yeah. Sure. Thanks to you. I I have okay. One thing. So I I noticed that the latest version is not on on GitHub. So could please upload it as soon as possible so we can have people contributing with issues and PRs. If you have any difficulties in handling the the Git workflow, just message me and and Michael, and then we will help.

[00:33:17] **Jungha Hong**: And then

[00:33:17] **Lorenzo Corneo**: I have also yeah? Sure. And then I I have another question on the previous slide, actually. When you flashed the idea of supporting AI agents, so I'm wondering what is needed. So in just just for completeness, at the last IETf in Vienna, Ari, myself, and Jaime had a draft on some extensions for, SDF and the CoAP for making these technologies ready for for AI agents. Mhmm. It was it it is in the thing to things, research group. I can link it later. And, I wanted to know what is it that you would like to have, because if there are some needed capabilities, then we we we need to consider, having a work covering those. And and and perhaps it's time that we discuss this in the working group and and and and get more concrete about that. That that's my question.

[00:34:33] **Hyunjeong Lee**: I think the digital twin needs AI based examples. So, firstly, I think that we have four examples for digital twins, so we we can add some AI agent components, things like this. So AI based components like the for the inference decision or control. So for the AI agent, I I just add some AI components for for each agent, things like that.

[00:35:11] **Lorenzo Corneo**: Okay. Okay. Thanks. So this is actually a bit different from what we have wrote in the in the thing to think, RG. By the way, I I sent the link here in the chat, so if you want to have a look at that. Mhmm. But but what we propose there is more for, yeah, embedding some instructions for AI agents into the SDF document while I think you're looking more for inference and and and control. So it's definitely a complimentary set of features that are needed.

[00:35:46] **Bart Brinckman**: Mhmm.

[00:35:48] **Lorenzo Corneo**: Yeah. I don't know if Ari has any input on this on on on the way forward or anyone else.

[00:35:59] **Ari Keränen**: No immediate feedback, but I think that's a it's a very interesting area and, of course, kinda interesting emerging domain. And, you know, happy to see you're you're looking at from the digital twin point of view. And and when you when are if I mean, of course, there's two parts to it. One is the current focus of the digital twin traffic. What how to model things using exit existing SDF features to support digital twins. And then the second part is, okay. Are there additional things needed? And the second part is, of course, the draft that Lorenzo mentioned. He's already, you know, discussing some potential. And then I think the the the what is very much complementary in this draft is, okay, how we take use of those capabilities in order to support the kind of scenarios you have in mind. But, yeah, looking forward to the, more deep detailed discussion on this topic.

[00:36:48] **Lorenzo Corneo**: Do we have any more question for Hyunjeong? Okay. Actually,

[00:37:00] **Hyunjeong Lee**: the AI AI based components just planned, so I'm not sure I can make the another draft for the AI AI based components. So maybe this is final or I plan to one or two more documents there for the working room last call. This is my plan.

[00:37:25] **Lorenzo Corneo**: Yes. Yes. Noted. Absolutely. It's getting there, I think, also the the draft. But if something for, AI agents is needed, then and and it has to be normative, then we need to have another draft addressing that so that you can incorporate it in your digital twin draft. So that that was my point. Thank you. Yeah. Sure. Thank you. But thank you, Hyunjeong and and Jungha as well. Great work. I guess we can move to Jungha with the non-affordances. Right? Okay. So I think control back. Thanks again, Hyunjeong. And now let me load the slides for Jungha. Gonna forward this. Jungha, I give you control of the slides. The floor is yours. Go ahead.

[00:38:33] **Jungha Hong**: Okay. Thank you. Can you hear me?

[00:38:37] **Lorenzo Corneo**: Yes. Loud and clear.

[00:38:40] **Jungha Hong**: Yeah. K. Okay. Let me start this. I'll give a short update on the our draft on SDF extension for a lot of further information. This slide, I'll briefly cover the changes in version actual four and five, and then I'd like to discuss some point for next revision. The version three was expired, so we submitted version four to keep the draft active. So there was no technical changes for that. So I skip this. And for budget five, we made some more substantial updates. First, we updated the base, a step reference to the RF c nine eight eight zero. We also added a former definition of SDF context. And and then we also aligned the, qualities with RFC 9 eight eight zero. And finally, we updated section four point one point three to align with the latest SDF-Type-Link draft. In particular, SDF-Type-Link is now used with the type object and href. Actually, this slide is for the discussion. I have some point to discuss in this meeting. So I go over to all the slides, and then at the end, I would like to get the feedback from you. So I would like to keep this draft focused on the the SDF context modeling only. Currently, this draft defines the three runtime messages like context snapshot and identity manifest and context patch. But but the the group has a instance information draft, which provides a general framework for as they have instant information, and it covers the instance values for both avoidance and non-affordancy. So we propose to keep this draft focused on the SDF context modeling only and handle the runtime representation and the instance information draft. So the the the the SDF context defines a nonaffordance information at this at the model level while the instance information handles its instance value at runtime. So we could we would like to propose to move the runtime representation to the instant information draft maybe in the next revision. And this is one suggestion. And if it makes if we do, then it the the changes will show like this like this picture that the at the SDF model, we have a affordance on the left. It has SDF property, SDF action, and x SDF event. On the right side, we have SDF context represent the nonaffordance information. And both types of the information can be can have the instance values. So at the, instance level, they can both be handled by the SDF instance information framework. So this gives us a clean separation between the modeling and instance information. And then the in this version, we updated sesh the the section four point one point three to align with the latest draft of the Sdf type link. But after reading it again, we think the most of this section describes the general semantics of the about the link. So the those semantics are not specific to the SDF context, and they are already defined in the other draft. So for the next revision, we propose to remove this as a separate subsection instead. We can keep it as a simple example to show how an external resource can be represented as nonaffordance information using SDF context. And then so this this is just plan. Based on these changes, this is our proposed structure for the next revision. So the main change is to make the scope of the document clearer. So section four will focus on the definition and semantics of SDF context, including where it can appear, its qualities, how it differs from SDF property, and processing requirements. And the use cases and examples will be separated more clearly. The external resource example will also be moved into the example section. We also propose a new section on on the relation ship to other SDF documents in class six. This can explain how this draft relate to the instant information and SDF-Type-Link and other related work. So the goal is to make this document more focused on the SDF context model extension. Yeah. Yeah. This is all I have today, and I would like to, get your feedback for my two, my suggestion for moving the runtime messages to into the instance draft. I think the I also wonder the instant draft has been expired. I wonder what is the planning for that. Thank you.

[00:46:29] **Lorenzo Corneo**: Thank you, Jungha. I actually like the proposed changes. This would make the draft more specific, more narrow. I if this is confirmed, I would also recommend that maybe we have a you can change the title, for example. I don't know. SDF context, colon, nonaffordance, something something. And, yeah, and I I also have the same question that you raised about the instance information, and we have Carsten in in in the call. Yeah. He he he just messaged. Yeah. I I I think he he agrees on on on moving that material to the instance information draft. So, yeah, I'm very positive on on on the changes that you propose.

[00:47:30] **Rohit Mohan**: Okay.

[00:47:33] **Lorenzo Corneo**: Absolutely.

[00:47:34] **Jungha Hong**: I'm glad to hear that. Thank you.

[00:47:36] **Lorenzo Corneo**: Yeah. So I

[00:47:41] **Jungha Hong**: I would like to hear from Carsten what is the plan for the instance draft. We can't actually actually, we can we can work. We can move the things from our draft to their draft to update it in their next revision.

[00:48:09] **Carsten Bormann**: Yeah. So, obviously, we we need to update the instance document. And if you can can give us material that you want to be covered in the instance draft, we can add that. So the this should happen before the the San Francisco meeting. And Mhmm. Yeah. The the

[00:48:42] **Lorenzo Corneo**: more

[00:48:45] **Carsten Bormann**: specific the information is you can give us, the the more we can make sure that this actually fits together well.

[00:48:54] **Jungha Hong**: Mhmm. Mhmm. Yeah. Okay. Thank you.

[00:49:07] **Lorenzo Corneo**: Do we have any more comments from from the audience? Yeah. Also, Ari's plus one. Yeah, Rohit?

[00:49:17] **Rohit Mohan**: Hey, RJ. Just a quick question. Do we have some kind of formal definition for the SDF context, like a CDDL product? I don't see it on the top.

[00:49:34] **Jungha Hong**: Yeah. I think in the current draft, I think it makes some confuse. There is some description instead of definition. So when we revise in the next revision, we as we propose the structure of the draft, we will add new section for the definition and semantics. So you will see in the next clearly in the in the next revision.

[00:50:10] **Hyunjeong Lee**: Thank you.

[00:50:17] **Lorenzo Corneo**: Yeah. Nice. Anything more? Okay. If if there are no more questions, I I want to thank Jungha for for the nice presentation and great work. And so this concludes our agenda. And then I'm so so now we are in the is there any other business to to discuss? Bart?

[00:50:55] **Bart Brinckman**: Yes. Just some just to let you know, so we we've also so we we have NIPC slash SDF protocol mapping sample app. It's one of the implementations. What we've also done recently is, extended that with a tool that, builds, like, Zigbee slash BLE SDF models from an actual device. So what it does, it reads the gap table or the Zigbee attribute table and automatically, you know, basically creates the the SDF model. So we we have a few SDF models that, you know, are being used by by, you know, kind of our customers slash partners, and now we have the ability to quickly generate more. So I'm thinking it becomes it's becoming time that we have a repository for our for our models as well as, you know, some of the open source tools we have. So I I think, for example, Carsten did a really good job with, you know, cboard.io, kind of putting all the resources together. So I've I've registered I was able to find sdf.wiki. So I'm thinking if we could create a in the in the GitHub, if we could create a repo, we could potentially collaborate on such a such a web page you know, makes this available to the community. I'm I'm not sure what the thoughts are on that.

[00:52:53] **Rohit Mohan**: Ari? Yeah.

[00:52:54] **Ari Keränen**: Thank you, Bart. I mean, this this is this is really good, something this actually has a spot on, like, what we also be discussing in the one DM context. So definitely, let's let's have a chat on on that how the best way to do it. I think we all share a common goal here. We could actually I wonder what what is the best for for that because that goes a bit beyond the typical ASDF work. So the one DM might be the right one DM would be the right maybe the right fora for this collecting them models and doing the publicity part of it. So could coordinate on having that discussion in a one DM call.

[00:53:38] **Bart Brinckman**: Oh, okay, Ari. I mean, if yeah. If we I'm I'm happy to participate there or so if you, you know, can point me to that, we can have the discussion there if you want. Yeah.

[00:53:52] **Ari Keränen**: Sounds very good. Yeah.

[00:53:55] **Lorenzo Corneo**: I I agree with the Ari. Cool. Let's do something about it. Yeah. Yeah. Maybe one DM is the right forum for that.

[00:54:07] **Bart Brinckman**: Okay. Yeah. You well, you you know how to find me. So

[00:54:12] **Lorenzo Corneo**: Yes. Yes.

[00:54:14] **Ari Keränen**: Yes. Yeah. Let let let's let's chat over Telegram to find a good time for that.

[00:54:19] **Rohit Mohan**: Okay.

[00:54:20] **Ari Keränen**: That's the that's the best option, I think.

[00:54:26] **Lorenzo Corneo**: Okay. Any other business? Jungha?

[00:54:33] **Jungha Hong**: Yes. I we got a email from the Michael, I guess, that the ASDF working room may be concluded soon. So I wonder what is the plan for working group. So depends on that. I think we we would like to plan for our draft team, the documents.

[00:55:12] **Lorenzo Corneo**: Yeah. I I'm not aware of this, Jungha. I can discuss that with with with Michael. Let me check my email. Yeah.

[00:55:27] **Jungha Hong**: Is there any is there a possibility that really the ASTM working group can be closed?

[00:55:39] **Lorenzo Corneo**: Eventually, yes. But I think we are not there yet.

[00:55:48] **Ari Keränen**: Sorry. I I missed for the notes. What was the original question?

[00:55:57] **Jungha Hong**: Oh, and I don't know when it was, but the I think Michael sent an email to the list that the the ASDF working group may be closed next next year, earlier.

[00:56:25] **Ari Keränen**: Okay. Now I don't recall such email, but I but, of course, like, once we're done, then the ASDF group would would be closed. But I guess we're not quite there yet. And and now it depends now it depends on the timeline with the rest of the work. But as as Lorenzo said, like, once we're done, of course, we're we're close, but I guess we still have a few documents that we have on the charter and and a plan to finish those.

[00:56:49] **Lorenzo Corneo**: Yes. Maybe we we can reconsider we can consider rechartering at some point even if we believe that all the milestones have been met, but we need more. That that is a possibility as well. So don't worry. Your your draft are safe.

[00:57:06] **Jungha Hong**: So Okay. Yeah. Okay. Okay. Thank you.

[00:57:10] **Lorenzo Corneo**: Yeah. Sure. No problem. Okay. Well, if there is no other business, we can even close it here. It was a very good meeting. We, yeah, I want to remind you also that we have already scheduled another interim before IETF one to seven. And as far as I know, there's not gonna be an ASDF meeting in San Francisco. But yeah. So these are my final words. And with this, I guess, I thank you all for participating and engaging in very fruitful discussions, And I'll see you roughly in in a month at the next ITF-asdf interim meeting. Thank you all. Have a nice day.

[00:58:04] **Hyunjeong Lee**: Thank you.

[00:58:05] **Jungha Hong**: Thank you. Bye.

[00:58:07] **Ari Keränen**: Bye bye. Bye. Bye.