<taki> scribeNick: taki
Sebastian: We have guests today.
... Guests from CG are W3C members.
<kaz> W3C Patent Policy
Sebastian gave information about patent policy etc.
<inserted> Nov-11
Sebastian: We had guests last
week.
... We checked TD 1.1 transition
... We discussed Thing Model.
... Thing Model, Device Model. Eclipse Vorto.
... We had guest from schaeffler.
... any objection to make minutes public?
... Kaz, please publish minutes.
... next TPAC meeting minutes.
... We discussed Thing Model. I presented slides.
... Slides are not linked.
... We checked status of TD 1.1 draft wrt Thing Model.
<kaz> vF2F Thing Model session
Sebastian: Discussed features such as
extensions.
... Discussed pull request #540.
... Thing Model's relationship to architecture document.
... any objections?
... no objections.
... Joint discussion wuth JSON-LD WG
<kaz> vF2F JSON-LD session
Sebastian: Discussed issue #988
... and #967
... Mainly discussed issues that we discovered during the
development of TD 1.1.
... Issue #643. JSON-LD and WoT approaches are somewhat
contradictory.
... Issue #988. Round trip issue.
... LD-proofs is about security.
... Discussion is still on-going.
... any objections?
Sebastian: transition is approved.
... Kaz found an issue.
... I will going to address this issue.
Kaz: the remaining issue is completely editorial. I can update both template and index.html.
Sebastian: I already introduced WoT to
the CG.
... This time, linked building data CG introduce us what they
are doing.
Georg: I have been working on
Building-related research.
... There are some members joining from CG today.
Lagally: I was in WoT group 2 years
ago.
... I authored building topology ontology, and worked in SSN
ontology.
Maria: I contributed ontology
model in WoT group before.
... contributed to SAREF.
Mads: I am working on architectural enginneering. I worked on topology ontology, then joined W3C. I am working on building platform.
<GeorgSchneider> Slides are here: https://docs.google.com/presentation/d/1aZosv92Mg59I1ErHQUyRqNlTMMcIW9gwtwioXbWLrd0/edit#slide=id.ga10f9409b8_2_56
<GeorgSchneider> readable
Herve: I am from organization IIS. Modling on building automation. also in Linked data community.
Georg making presentation.
Georg: prepared with inputs from CG
members.
... motivation. BIM. application of BIM. increasingly used.
data exchange between tools throughout lifecycle. virtual from
physical through lifecycle.
... model by architect. later, engineering model comes in.
during operation, things in building for actuation. data are
heterogeneous.
... geographical data is important in building. Also Weather
data.
... IFC4 already exists.
... there are shortcomings. File-based. not modular.
... not web-compliant.
... IFCS5 is currently worked on in a better direction.
... CG bring together experts, to manage data across building
life cycle. Both researchers and practitioners.
... 140+ members as of september 2020.
... 2019 August, there was a change in chairmanship.
Georg shows CG web page.
Georg: We have GitHub page.
... We have bot, lbd, etc. repositories.
... We have public ML. every other week, there is a
telecon.
... meny members are associated with other SDOs, such as ISO,
ETSI, bSDD, CEN TC 442, etc.
... there are collaborations between other W3C groups as
well.
... We maintain github repo.
... bot (building ontology)
... building ontology is being developed again and again.
... there are common relationships.
... we found many relationships between ontologies.
... We came up with initial building topology ontology.
... there is a BOT document published in 2020.
... bot:Zone is part of physical world. general concept from
DUL.
... zones are different from different perspective.
Architecture, fire, thermal, etc.
... bot:Element examples are chairs, windowsm Air conditiners.
constituent of construction entity.
... Chairs can be put into building.
... relationship between concepts. zone relates to element.
contains, adjacent, etc. relationship are defined.
... Matryoshka style nesting. building has multiple stories.
stpries has spaces.
... zones can intersect. two stories, elevator spans across
multiple stories.
... bot:interface is about surface.
... zone and wall. consider heat transfer area. interface is
between zone and wall.
... if there are pipe between zones, it can also be
interface.
... is it possible to map BOT to brick, ifcOWL4, SAREF4bldg
etc?
... BOT is upper level ontology.
... there is a draft community group report in github.
... Next about implementations.
... BOT + SOSA + Geometry were combined using Web interface.
There is a YouTube video showing this implementation.
... IFC-to-LBD converter. There is a GitHub repo. IFC has many
tools for integrating data. It outputs building topology
ontology data.
... Where are we heading at?
... CG is an active forum. building data on web best
practices.
... Linking to WoT TD. BOT elements should be able to link to
TD.
... Thank you. Any questions?
<Zakim> dape_, you wanted to What makes IFC5 more web compliant? more/other Industry Foundation Classes or also changes w.r.t. to format and such?
Daniel: IFC seems to be using a format.
Georg: They are moving.
... They are shifting from STEP to web-compliant format.
<GeorgSchneider> https://www.buildingsmart.org/about/technical-roadmap/
Daniel: Linking is not very possible at this point, I think.
Mads: Python library can help.
Sebastian: Do we have specification where ontology is hosted?
<GeorgSchneider> All ontologies are hosted here: https://github.com/w3c-lbd-cg/bot
Sebastian: Do you have a fixed URL?
<GeorgSchneider> https://w3id.org/bot
Georg: Yes.
<mlefranc> see https://github.com/w3c-lbd-cg/bot/issues/85
Sebastian: We can think of use cases about how TD can be used with BOT. We can work on a Note document together.
<mlefranc> we have issue https://github.com/w3c-lbd-cg/bot/issues/85 for a JSON-LD context that could be used in conjunction with the TD context
<kaz> scribenick: dape_
Cristiano: interested to work
together
... Question: Ontology is high level. Possible to describe
bridges?
... not being "actual" building
Mads: BOT is meant for
buildings
... one could use the same mechanisms
... Norway did something similar.... works exists
<GeorgSchneider> Definition of bot:Building: Building - An independent unit of the built environment with a characteristic spatial structure, intended to serve at least one function or user activity [ISO-12006].
Georg: definitions are specific to buildings
McCool: +1 for capturing use
case
... we do have repo for that
<inserted> wot-usecases repo
McCool: TD examples are very
interesting
... in PlugFest we used geolocation with links
... talking about zones / buildings sounds interesting
<McCool> https://github.com/w3c/wot/issues/939
McCool: spatial data on the web
meeting
... scenario: HVAC systems ... coordinate sensors with
actuators
... oneDataModel might be of interest also
... SSN talks about feature of interest
... one driving example might be useful
Georg: Question: is there a way to specify location in TD?
McCool: no standard way yet
Sebastian: TD model is meant to be
neutral
... location information highly depend on use cases
... that is why we did not pick one
... 1.1 version introduces how location based information can
be used... using existing ontologies
McCool: should facilitate convergence
Herve: MQTT: meta-data embedded in JSON-LD ?
McCool: TD is more meant for static
information
... dynamic data is more coming from properties
... We do support MQTT bindings
Sebastian: Future plan?
... becoming working group?
Georg: No
... keep community group as is
... spin-off working groups are possible
Sebastian: I see
Georg: For proper standard wg is
needed
... we plan to keep CG running
Max: Ted mentioned we do not need to create WG, but we can add topic to the newly proposed Spatial Data on the Web WG Charer
Sebastian: Suggest addressing topic in
use case task force
... e.g., creating W3C note
Georg: Sounds good. Support this effort
Sebastian: Will ping M. Lagally
... inviting you in the use case call
... new time slot is planned for the upcoming use case calls
... will let you know once the time/date is set
Georg: Great
Kaz: bringing this to the WoT Use Cases TF and generate some use case description would be great. On the other hand, we should think about the relationship between this ontology proposal and others like oneDM, schema.org
McCool: we have examples combining several ontologies
Kaz: e.g., mashup use cases?
Georg: Yes, there are lots of
ontologies
... link to the ones that are recognized
... propose best practices
Kaz: starting with use cases is correct. At some point we need to look at how to integrate best.
Sebastian: Okay, lets discuss this in use case meeting
<inserted> Issue 1000
Sebastian: Issue#1000
... Kevin shared Vorto example
... washer example
... contains properties readOnly, events (partial) and actions
(partial)
... "type" reference in properties
<cris> +1
Kevin: Yes, the current link points to complex type and we might need to fully define it
Ege: Datatype representable with JSON schema?
Kevin: Yes, should be possible
Ege: Where do I find the datatype in vorto lang?
<Ege_> ?
Kevin: either repository or same folder structure (in Vorto)
Ege: how is "type" used? Checks?
Kevin: Vorto lang tooling tries to
import and resolve
... similar to Java imports
Cristiano: Comment: We might need
another keyword ... express model of property
... type does not seem to be the right fit
Kevin: Note: Complex types are aggregrates of simple types
Cristiano: @type is probably not useful either
Sebastian: "mandatory" flag ?
Kevin: payload is similar
Ege: "required" keyword can be used
Sebastian: Let's discuss further in
issue
... issue 999, see https://github.com/w3c/wot-thing-description/issues/999
... ML is not here -> postpone
... issue 1001, see https://github.com/w3c/wot-thing-description/issues/1001
... about "precision"
... about "reliability" similar issue, https://github.com/w3c/wot-thing-description/issues/1001
... ML asked for additional terms
McCool: SSN defines already accuracy and precision
Ege: does not really define what it means.. I think
McCool: I think they do define
<Ege_> https://www.w3.org/TR/vocab-ssn/#SSNSYSTEMAccuracy
McCool: about "reliability" i am not sure if ontology exists
<inserted> Issue 1002
Cristiano: should point to other
ontologies OR add another layer
... SOSA layer and SSN
Ege: I do not see "reliability"... but we need to be sure what ML means by that
McCool: Could be accuracy... but I
am not sure
... false reading falls into accuracy
Sebastian: I am bit concerned. This
information depends a lot on environment / use-case
... not sure if we want to introduce these terms
McCool: We can recommend terms
Cristiano: Should we deprecate unit? Why do we have unit in the first place
McCool: Broader concept... and we have it already
Sebastian: motivation for unit was to
provide a simple solution for JSON solutions only..
... not necessarily based on ontology
Cristiano: Not sure were to put the line
Sebastian: Agree. We should be
careful
... "stability" was used in the past also
... not sure about the benefits
McCool: stability was meant for
support caching... that is, it had an operational purpose
... reliability might be different although might have an
operational meaning for analytics (eg to indicate that outliers
should be discarded)
<inserted> wot-binding-templates issue 103
Sebastian: issue#103
... how to propose a new binding
... worked on PR also
... Another proposal: Restructure document
... maybe we can have a dedicated section for each protocol,
with subsections etc
... at the moment the information is somewhat distributed
... OR one document for defining basic concepts and
side-documents for each binding like HTTP, CoAP, ...
Cristiano: +1 for revising structure
Ege: multi documents advantages would be multiple editors for each binding
Daniel: +1 for Ege ;-)
Sebastian: Will create issue
... let's continue discussion on GitHub or next week
<kaz> [adjourned]