Scott: background on discussions in ISG DWG, Smart Cities DWG, Metaverse ad hoc all had significant overlap
Chris Little: virtual world that reflects the real world; modeling what "is there" and what "could be" - some things may not be geospatial
Josh: there is a lot of spatial thinkingin our social dimensions. Digital Twins are a unifying concept.Metaverse may also include"Digital Distant Cousin"
Nadine: take aways from Metaverse session:
Explaining value of spatial to outsiders; some of this is ebing reinvented
These topics are not just for OGC, need to link to other groups working in the space
need to demonstrate what we are doing
Carsten: there are almost too many overlaps amongst all of OGC's Working Groups. Have not realized as much as maybe we should have from Digital Twins Location Powers event. A lot of work on digital twins in the built environment in the UK - includes an ethics framework
Kyoung-Sook:https://www.digitaltwinconsortium.org/initiatives/the-definition-of-a-digital-twin.htm
Josh:https://oascities.org/
Mark Korver:https://www.geospatial.jp/ckan/dataset/plateau-tokyo23ku
Mark Korver: the value of large 3D datasets like OpenStreetMap and collaboration to build content
David Graham: have a "need for speed" - lots of commercial creation of these datasets[synthetic environments...something 'more' than datasets]
Josh: the issues of digital "real estate" are very real and very open right now.
Chuck: we have 3D Standards today, we have the ones we need (to build digital twins, metaverse)
Dean: MUDDI discussions - the digital twin could be represented from a family of Standards. OGC could use a Best Practice to tell how to use these Standards
Kyoung-Sook: Recently I heard ISO, IEC, JTC1 mentioned city information model, platform, data sharing, etc for smart cities. They talk about the geospatial but not looking at OGC. Josh: the spatial component is so inherent in their thinking that they don't even consider it a distinct area of expertise.
Carsten: the 3D elements are not always available, but the 2D is still useful (fit for purpose)
Dean: a dominant theme this week has been the value of metadata, particularly live metadata that is constantly updated
Chris: metadata is often dynamic - can be updated more frequently than the data!
Nadine: crossing domains requires semantic linking
Stan: with respect to Smart Cities, need to consider autonomy - the ability to use all of the data almost immediately. Be a smart city, not just an informed city
Chris: problems of scale and speed
Brent: is OGC considering a global conflation framework? Data across scales, leverage overlap. Cannot have a human-in-the-loop with autonomous systems
Chuck: key is conflation in context
Brent: how to we enable machine learning on spatio-temporal data, need to embrace all of the data
Scott: the digital twin still must be created in the context of how the twin is to be used
Chuck: a model emulates the real world
David: the term "twin" is problematic - we don't need an atom-for-atom replicant
Chris: twins/models may be ill-conditioned and deviate from reality unexpectedly - "chaotic behaviour" - constraints/conditioning needs to be well understood
Stan: For Smart Cities, we may need to consider Digital Twins from the perspective of a set of Digital twins. For example, with data pulled in from sensors.
Some digital twins are just a construct of a workflow and may not need a human to see it
Dean: need well-defined use cases to tell us why we need the digital twin (or what we need it for)
Carsten: there is an important societal aspect; ethical concerns must be monitored. Not just a technical domain.
Are the people living in a city that is modeled as a digital twin part of the digital twin?
Privacy, security - look at Data Centric Security ERs (Testbed 17, Disaster Pilot)
Privacy and security should be build in from the start - privacy by design; not be patched on later
Josh: autonomy requires trust/accountability
Chris: exmaple of variable speed limits
Dean: low-hanging fruit - what is the minimal set of information needed? What Standards should be used
Dean: possible minimal twin data metadata model? resource uri,url, standard/protocol, reference frame(xyzt min-max) - anyone have references related to work done on this re dig twin context?
Generating more and more content that is predictive, not observations
Chris: predictions can be part of the boundary of a model
Gobe: Semantic similarity approached to navigate knowledge graph
Discussion in Closing Plenary on 2022-03-03
Scott: background on discussions in ISG DWG, Smart Cities DWG, Metaverse ad hoc all had significant overlap Chris Little: virtual world that reflects the real world; modeling what "is there" and what "could be" - some things may not be geospatial Josh: there is a lot of spatial thinkingin our social dimensions. Digital Twins are a unifying concept.Metaverse may also include"Digital Distant Cousin" Nadine: take aways from Metaverse session:
Carsten: there are almost too many overlaps amongst all of OGC's Working Groups. Have not realized as much as maybe we should have from Digital Twins Location Powers event. A lot of work on digital twins in the built environment in the UK - includes an ethics framework Kyoung-Sook:https://www.digitaltwinconsortium.org/initiatives/the-definition-of-a-digital-twin.htm Josh:https://oascities.org/ Mark Korver:https://www.geospatial.jp/ckan/dataset/plateau-tokyo23ku Mark Korver: the value of large 3D datasets like OpenStreetMap and collaboration to build content David Graham: have a "need for speed" - lots of commercial creation of these datasets[synthetic environments...something 'more' than datasets] Josh: the issues of digital "real estate" are very real and very open right now. Chuck: we have 3D Standards today, we have the ones we need (to build digital twins, metaverse) Dean: MUDDI discussions - the digital twin could be represented from a family of Standards. OGC could use a Best Practice to tell how to use these Standards Kyoung-Sook: Recently I heard ISO, IEC, JTC1 mentioned city information model, platform, data sharing, etc for smart cities. They talk about the geospatial but not looking at OGC. Josh: the spatial component is so inherent in their thinking that they don't even consider it a distinct area of expertise. Carsten: the 3D elements are not always available, but the 2D is still useful (fit for purpose) Dean: a dominant theme this week has been the value of metadata, particularly live metadata that is constantly updated Chris: metadata is often dynamic - can be updated more frequently than the data! Nadine: crossing domains requires semantic linking Stan: with respect to Smart Cities, need to consider autonomy - the ability to use all of the data almost immediately. Be a smart city, not just an informed city Chris: problems of scale and speed Brent: is OGC considering a global conflation framework? Data across scales, leverage overlap. Cannot have a human-in-the-loop with autonomous systems Chuck: key is conflation in context Brent: how to we enable machine learning on spatio-temporal data, need to embrace all of the data Scott: the digital twin still must be created in the context of how the twin is to be used Chuck: a model emulates the real world David: the term "twin" is problematic - we don't need an atom-for-atom replicant Chris: twins/models may be ill-conditioned and deviate from reality unexpectedly - "chaotic behaviour" - constraints/conditioning needs to be well understood Stan: For Smart Cities, we may need to consider Digital Twins from the perspective of a set of Digital twins. For example, with data pulled in from sensors. Some digital twins are just a construct of a workflow and may not need a human to see it Dean: need well-defined use cases to tell us why we need the digital twin (or what we need it for) Carsten: there is an important societal aspect; ethical concerns must be monitored. Not just a technical domain. Are the people living in a city that is modeled as a digital twin part of the digital twin? Privacy, security - look at Data Centric Security ERs (Testbed 17, Disaster Pilot) Privacy and security should be build in from the start - privacy by design; not be patched on later Josh: autonomy requires trust/accountability Chris: exmaple of variable speed limits Dean: low-hanging fruit - what is the minimal set of information needed? What Standards should be used Dean: possible minimal twin data metadata model? resource uri,url, standard/protocol, reference frame(xyzt min-max) - anyone have references related to work done on this re dig twin context? Generating more and more content that is predictive, not observations Chris: predictions can be part of the boundary of a model Gobe: Semantic similarity approached to navigate knowledge graph