[David McKee]

Thanks @Tom Hughes for hosting, I’ll keep an eye and respond to further questions periodically as well. And of course the invitation remains open for anyone who would like to explore a case study to try and practially explore some of these topics

[David McKee]

A really good point that we’ve started looking at. There’s an example of a company in the US that was recently acquired by an insurer who was trying to do this but with much less data than we’re talking about here: https://www.insurity.com/solutions/insurity-data-analytics-solutions/spatialkey-solutions/

Again this could be a great use case that spans multiple domains and I’d be interested in brining together 2 or 3 stakeholders around a case study to see what’s possible

1 minute ago, Guest Katharine van Someren said:

Hi David, 
Thank you for your responses.
Have any specialists connected with insurers to discuss any possibility of reductions in insurance premiums for the built environment where digitally enabled infrastructure is present?  The background behind my reasoning would be related to picking up changes before major or critical failure.  
 

 

[David McKee]

Thanks @Tom Hughes we’ve just finished out Digital Twin series and we’re actually about to start an article series on data formats, sources etc. It’s also worth looking at the Digital Twin Consortium website: https://www.digitaltwinconsortium.org/index.htm

[David McKee]

1 minute ago, Steven Zhang said:

Hi @David McKee 

Thanks for answering the question, I’d love to follow up. I have been using relational database, to manage some less real-time data, and also used graph database to explroe the relationship of meta data.

The use case I hope to provide is to use some of the more and more open geospatial sets, to provide a context for digital twin. Basically, it provide a interconnected spatial-data identifiers, that can be used to integrate/aggregate a range of data at various spatial granularity. It is a national data framework, and the spatial identifiers are commonly used by various oganisations, for administrative purpose and for statistical purposes.

[David McKee]

1 minute ago, Guest Katharine van Someren said:

Hi Tom & David,
Another really interesting and thought provoking presentation, thank you.
David:
Where in your work has the simulated digital twin helped decision makers explore and understand the connections between infrastructure?
Looking at scenario development – do you please have examples of where you have looked at built environment disaster recovery for power failure or flood risks?
Many thanks,
Katharine
 

Answering part 1, in Leeds and another project in the midlands once we’ve mapped the various infrstructure datasets/or IoT sensors into a common geospatial format we can trace through network behaviour the causal relationships. For example:

• Satellite imagery from ESA Sentinel programme gives mm accuracy ground movement every 6 days
• At the same long/lat there might be reports of a water pipe leakage
• These can be correlated to imply a relationship but requires a domain expert to either check by visting the location or doing further analysis

We haven’t yet done anything around disaster recovery but would love to have a chat about it and see if we could do a case study with you.

[David McKee]

Just now, Steven Zhang said:

Hi, @David McKee  Before contributing a use case. Frist of all, I loved your talk, I loved your introduction of a cloud environment for digital twin that will be available for everyone. 

I noticed you are using a “unified graph-based structure” – that underpins common geospatial reference, does it mean a graph database, do you mind expanding on this? What technology are you using? How scalable is it? 

Very good question, we still ponder over this once a week! The tradeoff is graph datastructre vs timeseries datastructure and the sheer scale of the data that can be generated and may need to be processed in real-time. Our experience is to use a mroe traditional horizontally scaling SQL database with the support of tools such as Apache Spark. This also helps with interoperability using the likes of Apache Pig.

Happy to dive into more detail.

[David McKee]

1 minute ago, Tom Hughes said:

When it comes to interconnecting digital twins across organisation and sectors within the built environment what do you see at the key challenges and opportunities?

 

Almost as I mentioned previously:

• Data formats, could we identify the most valuable types
• APIs, do we want to support OpanAPI 3.0, GraphQL, IEEE HLA, or something else
• And then understanding the costs and business models behind it

Hopefully a set of case studies would allow us to practically find this out

[David McKee]

To build on this discussion, a call-to-action from me would be could we identify 1 or 2 case studies from the community where you would like us to try and build a digital twin. We can accept that we will probably get it wrong to begin with but the learning process will be worth sharing in terms of the: platform stack that is required to build them, the data formats and APIs, and even the user experience that is required to make it useful.

[David McKee]

Just now, Tom Hughes said:

To better understand maturity toward digital twins and the National Digital Twin programme CDBB have recently launched an extension to the Smart Infrastructure Index. This is currently open to DT Hub members from asset owning organisations until 31st August.

@David McKee – How have you seen maturity towards digital twins develop in the built environment? In your opinion, in what areas would increased maturity accelerate the delivery of the National Digital Twin?

The quick answer, is I think it’s slightly too early to tell. The key problem we’re seeing is the wide number of standards that exist in data formats, communication protocols etc. And Digital Twins have in reality been lead by tech company’s who until very recently didn’t engage with experts from the built environment domain.

So a short answer is we need key experts from BIM and related concepts to work with technologists and come up with almost a feature matrix of what needs to be supported and how.

[David McKee]

To add to that quick response, at Slingshot we’ve been working with Leeds City Council, Mott MacDonald, and the Open Data Institute over the last few months to build up a Digital Twin of Leeds City Centre that will be incorporating vehicle emissions and the impact on respiratory health of commuters as part of a transport planning exercise.

[David McKee]

@Tom Hughes I the most obvious starting point is cities, probably with a focus on transport and mobility. This builds a basis for extending further into built environment as well as environmental concerns such as vehicle pollution and climate change.