[Kirsten Lamb]

Here are some challenges that our researchers have brought up in developing digital twins, paraphrased by me, so if they are in error the fault is mine and I welcome corrections:

• The value of digital twins in providing the right information at the right time, so a key challenge is determining the frequency and timeliness of data collection to provide useful, valuable insights to asset owners.
• With satellites, InSAR and other earth observation technologies, a challenge is in processing the high volume of data needed to quality-check the measurements taken in a timely manner.
• In creating a digital twin of a building, existing asset management processes have been established to take advantage of the knowledge of human asset managers and the data provided by building management systems. A key challenge is to develop digital twins that are capable of complementing these existing sources of knowledge and data by adding new value.
• Computer vision can help identify events and behaviours in the built environment without capturing footage of people, making it more acceptable from a privacy perspective. One important challenge to address is giving machine learning algorithms a fully representative training dataset so that biases are not introduced into the resulting data.
• Each sensor in a building or asset may only be able to detect one factor or phenomenon in isolation, but if multiple sensors become networked together in ‘smart’ ways, they may be able to detect ‘complex events’, events characterised by multiple phenomena happening in a specific order, time frame or physical orientation. Understanding how to combine sensors and human understanding into truly ‘smart’ buildings that can detect complex events and respond appropriately is a challenge.
• One promise of connected digital twins is seamless services provided to the public through digital technologies in the built environment. When designing a comprehensive service ecosystem enabled by connected digital twins, it is difficult to break down existing siloes: from a technical data sharing and interoperability standpoint; from a regulatory and geographical standpoint; and from the standpoint of existing processes and business models.
• When designing services based on connected digital twins, it is important to acknowledge the inequalities in access to digital technology based on socio-economic, geographic, age, education, ability and other factors. Exclusion from services or inequality of service provision based on these factors is a major issue to consider in the governance and development of connected digital twins for the public good.

[Kirsten Lamb]

I’m attaching here my slides from the 24 August Gemini Call introducing the Digital Twin Journeys project.

[Kirsten Lamb]

Seren, I’m just going to drop a couple of links to projects that integrate environmental data with other sources of data, and were flagged to me while I was sourcing information and participants for the workshop:

• Flood Damage Economic Model: https://www.eco-x.tech/portfolios/fdem/ 
• HELIX: https://www.eco-x.tech/portfolios/helix/ (I believe this is what Anglia Water are using. There should be more information elsewhere about this project on the DT Hub site if you look through the case studies and other threads.)
• The Environment Agency uses data from various sources to understand the impact of interventions on flood risk reduction (or at least reducing the risk of damage). They’ll have published some of that information online in reports, but I don’t have any links to hand.

[Kirsten Lamb]

On 15/06/2021 at 14:27, Seren P said:

Did the report summarising the insights from the workshop spoken about above get posted anywhere?

Hi, Seren! It’s due out next week. Thanks for your interest! I agree with Henry, the space where this is most advanced is in water utilities, and also flood modelling. The Environment Agency, especially Karen Allford, would be a great resource for you on this.

Best wishes,

[Kirsten Lamb]

I’ll start us off. There was certainly interest in continuing collaboration between built environment and natural environment modellers based on feedback in the workshop. In the short term, what do we need to do to demonstrate the value of integrating data and models between these sectors? What cases of collaboration already exist? Flood modelling springs to mind, but comment if you know of others! Or, if none spring to mind, what would be a good place to start integrating these models? What local-scale projects would benefit from model integration?