– DRAFT –
WoT Use Cases

document status

McCool: we should check hyperlinks and identifies if they are references

Lagally: ok should I assign you to this task?

McCool: fine
… it is a clean up activity
… not super critical

<kaz> Issue 178 - Check hyperlinks and convert to references if necessary

<kaz> PR 175 - Update Health Privacy Section

Lagally: there's a minor issue in PR 175

McCool: not worry about it I fixed the reference to the real standard in other PR

Lagally: good work references seems good
… I think if there're smaller issues we can update later

<kaz> PR 177 - Update References

Lagally: line 8582 contains changes what have you done there

McCool: I re-organized the references
… and split them in categories
… added also comments

Lagally: no conflicts right?

McCool: no

Lagally: ok I propose merging
… merged

Guests

https://www.w3.org/Consortium/Patent-Policy-20170801/ W3C Patent Policy

<kaz> (WoT guys introduce themselves)

Lagally: Oracle, working on cloud technologies

McCool: co-chair of wot WG

Kaz: w3c team contact

Cristiano: IE, unibo research fellow working on WoT and IoT

Carlo: research fellow University of Pisa invited for presenting a conference research work

Mizushima: co-editor of use-case document

Virdis: Associate professor of University of Pisa

Christine: Interoperability and geospatial expert

Lorenzo: Phd student, interested in wot and blockchain technologies

Carlo's presentation

<kaz> -> slides tbd@@@

Carlo: the work was done in the context of the Arrowhead tool project
… we used standard WoT technologies: servients and scripting apis
… today I present you an use case in edge computing
… in the use case we have a patient with health sensors
… there are two software agents: a servient hosted in wearable devices and one in a gateway
… when the patient moves to the hospital the servient hosted on the gateway should move to a new node in the hospital
… where it can preserve its quality of service requirements
… now the problems are: state managment and addressing
… stateful migration can be transparent to the final wot application (preferable) or explicit
… when the servent migrate addresses need also to be updated

Carlo showing a UML flow diagram about migration process

Carlo: when the exposer detect a migration it emit an event to the consumers
… on the other end if the consumer detects an ip change re observe/subscribe
… to all the affordances

McCool: one note: webhooks and other protocols have the problem that when the ip changes the connection is broken

Carlo showing the expiremental setup

<McCool> (still have some followup comment, pls leave me on the q)

Carlo: there are some executing time costs for the consumer side
… we also measured the container downtime
… most of the downtime is due to the container migration time

Carlo: that's was all

Lagally: thank you, very interesting
… did you see any gaps in the specification?

Carlo: addressing might be a problem
… in wot adresses are considered to be static
… there are might be some security concerns when you migrate a container, how should those be handled?

Lagally: thank you, we can talk about it offline

McCool: follow up, we are working on this problem also in the network working group
… I agree to have a follow up call

Lagally: what about an use case description ?

Cristiano: how are you able to sent the onMigration event?

Carlo: we are using a reactive approach
… the servient detects a migration

and emits an event
… we are using coap

Cristiano: thank you got it

Lagally: the conversation is running long
… I think we should move to another call

<Zakim> kaz, you wanted to discuss state management (later) and to ask about the concrete scenario and device/network setting (e.g., HTTP, long polling, websocket, ...)

<McCool> (time check, we need to move onto Christine's topic)

Kaz: state management can be done in SCXML-based Interaction Manager, but we should have further discussion next time./

<kaz> SCXML

<kaz> MMI Architecture

Kaz: plus a detailed use case scenarios should be described.
… like data model, flow and protocols

Geospatial community

Christine: there's a new standard wg in the IEEE
… they are working in the spatial web
… on December they approved the first internal draft

<McCool> https://sagroups.ieee.org/2874/

Christine: chrstine: it is highly relvevant to connected things
… the second item is the open geospatial consortium
… THE most popular work in the geospatial communty come from OGC

<kaz> IEEE P2874 Spatial Web, Architecture and Governance Working Group|

Christine: on relevant work is geo posting
… it contains information also about orientation
… it uses also time
… the spec is trying to prevent the creation of different formats for indicating an orientation (i.e. in the context of building enviroments/ smart vehicles)

Lagally: I read the standard, a follow up meeting would be good about this topic

<cperey> https://github.com/opengeospatial/GeoPose/blob/main/standard/pdf/geopose_standard.pdf

Lagally: we need more time to collect ides and how it can be used inside wot

<kaz> i/geo psting/geo posing/

<kaz> s/geo posing/geopose/

Lagally: in general geo posing does not really fit all our use cases
… we need to integrate with other solutions

Christine: ogc would be glad to ear any limits of the standard

<cperey> IEEE P2874 WG 2:55 PM https://sagroups.ieee.org/2874/

Christine: and understand how it can be integrated with other solutions

Lagally: we are at the hour, can we continue in the next use case call

<cperey> Thank you for letting me know when to have an interim meeting, one week from today, I can also invite others

McCool: good

Lagally: aob ?

adjourned