Digital twins for buildings and infrastructure

Councils, utilities and estates teams look after buildings, networks and machines that have to keep running. A digital twin is one way of keeping watch on them. It is a working copy of the real thing, held on a computer and kept up to date with data.

This is a general guide. It is not AI Twin Limited and is not connected with it.

What a digital twin is

The Centre for Digital Built Britain describes a digital twin as "a realistic digital representation of something physical". What sets it apart from any other digital model, the Centre says, is its connection to the physical thing it copies. A twin can follow an asset through its whole life, from design and building to upkeep and the day it is taken out of use.

The data often comes from sensors fitted to the asset. They record how it is performing, its condition and its surroundings, and send that back to the virtual copy.

Where AI comes in

IBM lists an analytics engine as one part of a twin, "often powered by machine learning or artificial intelligence". It can spot patterns in the data, simulate what might happen next and make predictions.

One use is predictive maintenance. IBM describes it as teams anticipating problems before they happen, which can cut downtime and make an asset last longer.

The UK's national programme

The National Digital Twin Programme began in 2018, in response to the National Infrastructure Commission's report "Data for the Public Good". The Department for Business and Trade manages it on behalf of the UK Government.

The Digital Twin Hub, an online community for people working on twins, says a key purpose of the programme is to develop standards, frameworks, guidelines, methods and tools for twins. The programme's aim, in the Hub's words, is twins that are safe, secure and trustworthy, as well as ethical, both alone and when connected.

The Gemini Principles

The Centre for Digital Built Britain published the Gemini Principles in December 2018. They set out values to guide the development and use of the national digital twin. At the heart of those values, the Centre says, is the idea that "all digital twins must have clear purpose, must be trustworthy and must function effectively".

The Centre finished its five-year mission and closed at the end of September 2022. Its website stays up as a record of that work.

Securing the AI behind a twin

Security for the AI part of a twin is covered in the National Cyber Security Centre's (NCSC) guidelines for secure AI system development, written for the organisations that provide AI systems.

Useful links