-
Demographic Twins for ‘What if?’ Scenario Planning
Join us for the next video in our series on Tuesday. Mark Birkin and the CDBB team will host a live chat session at 10.30. Bring your questions.
-
20 Replies
-
Welcome to the start of today’s Digital Twin Talk on Demographic Twins for ‘What if?’ Scenario Planning and a big thank you to @Mark Birkin from the Alan Turing Institute for joining us. We’re looking forward to your thoughts and questions related to Mark’s talk – and maybe posing one or two of our own.
As with all our Twin Talks Mark is online from 10:30am to 11:30am to answer your questions. To join the discussion please add your thoughts by replying to the conversation thread.
-
@Mark Birkin you mention that demographic digital twins enable an exciting programme of academic research with extreme relevance to policy and real-world deployment. As digital twins of both national infrastructure and place based digital twins mature what do you see at the key opportunities for gaining further insights in demographic interaction, change and using data for the delivery of public good.
-
Real-time analytics would be close to top of my list – being able to access data on mobility or consumption patterns and predict and analyse trends and impacts is new and very powerful. we are starting to see some of this e.g. in transport and mobility planning – but I expect this to broaden significantly – in pubic health for example!
-
4 minutes ago, Richard Bradley said:
Hi Tom, what do I do – simply post questions here? Thanks Richard
Hi Richard, yes. You can post questions in this conversation thread. Mark, myself and other members of the CDBB team are online between 10:30 and 11:30 to answer them.
-
The ‘levelling up’ agenda is another area of significant opportunity. Organisations like the GLA and TfL at the moment seem to have a lot of resources and capability which I would like and expect to become much more widely available across metropolitan authorities and other local administrations around the country. For example we are working with colleagues in Bradford at the moment on a wide range of issues relating to inequality, ranging from diet and obesity, education, crime, and economic inequalities.
-
There is a nice introductory piece about the work in Bradford here:
https://lida.leeds.ac.uk/news/improving-lives-through-big-data-holly-clarke/
-
Thanks Richard, that’s an interesting question and I must admit I have a tendency to be suspicious of overambitious approaches (boiling the ocean”?!) but very much take your point about a level playing field. We do have a very open approach to the science of DTs – for example, outputs from the SPENSER model I discuss are already available through the CDRC at Leeds (www.cdrc.ac.uk) and all the covid models and code we are developing are open-sourced. Definitely something the Turing would like to build upon in the next phase of Digital Twin- I’m sure CDBB would feel the same way.
-
5 minutes ago, Richard Bradley said:
Hi Mark, at Transport for the North we were given the job to disrupt the analytical status quo and understand how investment decisions can be made more equitable. We have developed the Analytical Framework to achieve this, supported by evidence on why decisions may appear unfair. This applies just as much within the North as within the UK. The three themes that appeared from the associated ‘Five Whys’ root cause analysis included: exploring new futures and uncertainty; being able to show that levelling-up is good for the UK; and better representation of the customer experience. So a lot has been done over the past few years on levelling-up – I’m happy to take you through this experience if you want. Thanks Richard
That sounds good. I would love to know more. Also, any experience in translating these approaches to investment in other sectors e.g. housing, health care, education, crime – I guess you capture a lot of this through some kind of impact analysis?
-
Great presentation Mark! I like the way you describe population based Digital Twins. Infrastructure based digital twins will become more insightful if linked to population based models as you demonstrate. Currently the synthetic population twins use survey data and the development in real time analytics will bring these models closer to the reality of population Digital Twins. What are the challenges in getting hold of and sharing survey data? And what do you foresee as issues in verifying and sharing real-time data?
-
4 minutes ago, Mark Birkin said:
Thanks Richard, that’s an interesting question and I must admit I have a tendency to be suspicious of overambitious approaches (boiling the ocean”?!) but very much take your point about a level playing field. We do have a very open approach to the science of DTs – for example, outputs from the SPENSER model I discuss are already available through the CDRC at Leeds (www.cdrc.ac.uk) and all the covid models and code we are developing are open-sourced. Definitely something the Turing would like to build upon in the next phase of Digital Twin- I’m sure CDBB would feel the same way.
Yes – The trust layer of the Gemini Principles comprises of Security, Openness and Quality. This doesn’t remove or reduce innovation and differentiation, my view is a “level playing field” is an essential element for having meaningful competition and driving innovation.
-
There is a lot of talk about ethics and privacy of data, and these questions are clearly fundamental (great need to engage the public in understanding the value-added and benefits of using data in this way). But I’d see the main obstacles around ownership and commercial considerations. I think we can see this quite clearly in the current pandemic – there are great repositories of data that could dramatically illuminate our understanding of what’s happening and what to do within e.g. financial transactions, mobile telephone traces, but great sensitivity amongst the data owners to the terms under which these data might be shared for public benefit. More groundwork is needed – we can’t solve all these problems in the midst of a pandemic (though of course have started with institutes like Ada Lovelace – Royal Society also very active in this space).
-
The recent Cambridge Centre for Smart Infrastructure and Construction (CSIC) and CDBB publication Flourishing Systems – Re-envisioning infrastructure as a platform for human flourishing puts a renewed focus on; People, Connections, Sustainability and Digitalisation. My view is the kind of demographic digital twins you presented in your talk are critical for this human centred approach to infrastructure. In your view how can owners of infrastructure assets benefit from demographic digital twin and in your experience what is necessary to interconnect infrastructure digital twins with demographic digital twins.
-
The most obvious issue for me would be that infrastructure owners tend to have a concern with a population of users or customers, but the infrastructure impacts much more widely (e.g. through externalities like congestion and pollution) – so crucial that all our approaches (SPENSER, QUANT, MISTRAL etc) are grounded in understanding trends, patterns and impacts across the whole population. This underscores the importance of having a NATIONAL digital twin and I guess a need to balance interests of infrastructure providers with local agencies and some coordination through central government/ NIC and the like. The Turing I think is well-placed to tread a middle ground between these various interest groups, and also particularly well-placed to engage with government.
-
An open question to all in the discussion. What questions do you have for Mark about the Alan Truing Institutes work, or work elsewhere, on Demographic Digital Twins?
-
5 minutes ago, Richard Bradley said:
Hi Mark, the welfare benefits are largely captured through travel cost savings, which, in a perfect economy, includes capturing associated secondary benefits. However, there are wider benefits associated with economic failures in an imperfect economy and with agglomeration effects. These types of wider benefits are now largely captured. However, what is not as clear is the distributional effects and how these should be accounted for in decision making. For example, if we target infrastructure at a poorer or wealthier area we have to displace the impacts in some way as there is assumed a fixed amount of ‘water in the economy pipework’. However, this doesn’t level-up the UK and if we can understand quality of life then we might be able to understand distributional impacts. For example, the poorer area outcome could be higher wages, which then reduces crime, health issues, etc. This is tackling the ’causes of causes’ and goes across sectors. But we can’t do this effectively unless we can bring the silos together. DG’s have great potential to do this. Thanks Richard
Sounds good. And the microsimulation approach is perfectly suited to capturing these kinds of distributional consequences at quite a fine scale. The origins of these methods grounded in understanding consequences of taxation, benefits or financial policy across sub-groups, while advances in spatial MSM means that we can translate this into specific regional and local environments.
Log in to reply.