As the old saying goes, the map is not the territory. There is a gap between reality and depiction. Any map or representation of reality requires a level of abstraction.
Surrealist painter Rene Magritte succinctly made this point in his work The Treachery of images, which shows a tobacco pipe with the caption, “Ceci n’est pas un pipe” (This is not a pipe).
We cannot reliably infer the true nature of a territory by looking only at a map of it. But as digital technology continues to expand the horizon of what is possible, new solutions are emerging – representations that are not maps, but digital twins.
How did digital twins evolve? And what’s next? Let’s look through the lens of industry.
Historical precursors
Using the duplicate or ‘twin’ of an object to enable simulations and predict outcomes from operative changes, originated at NASA in the early 1970s. A physical replica of the imperilled Apollo 13 was used to simulate the damaged spacecraft. It helped to identify the possible causes of failure and potential solutions for the safe return of the crew.
The term ‘digital twin’ has been used since 2002. Michael Grieves applied it to product lifecycle management in manufacturing. Since then, the concept has evolved and been applied to many other areas. And its scope has increased tremendously.
From digital map to territory
In the last 20 years, we have moved from inert digital replicas to living and sensing simulations.
Today, with the help of machine learning, AI and statistical models, digital twins connect the real world by collecting real-time data from sensors and triggering actuators. Prompting humans to sit up, listen and take action, especially in industrial processes.
What does ‘digital twin’ mean today?
A digital twin is a crucial tool to help engineers and manufacturers understand how products are performing now, and how they will perform in the future. Digital twins give fresh insights and reduce uncertainty, which allows engineers to perform preventative maintenance. For example on large buildings, machines or vehicles.
Prevention, as they say, is always better than cure. In industry, ‘better’ can mean ‘cheaper’, ‘safer’ and ‘more efficient’ in various ways.
Digital twins in the real world
A striking example in industry is with Elon Musk’s Tesla. The electric car giant creates a digital twin of each vehicle it sells. Sensors from every individual car continuously stream data back to the factory. AI then analyses the data and determines whether a car is working smoothly, or if it needs maintenance.
For many issues, Tesla can release software or configuration updates that can fix or temporarily alleviate problems. For example, adjusting the hydraulics to compensate for a rattling door. By using digital twins and learning from the real world, Tesla can optimize each of its cars individually and in real-time.
Examples from Imperial College London
A recent example from Imperial College London’s Professor Mark Girolami, involved the realisation of the first 3D printed steel bridge in Amsterdam.
The performance properties of the materials were not fully understood. Through characterising and environmental sensors attached to the structure, it was possible to develop a digital twin of the bridge. This allowed the study of its evolving properties and assess what would be the effect of random forces of pedestrians, reducing the uncertainties on the project.
You can find out more about Dr Girolami’s work at his talk from Tech Foresight 2038:
Professor Girolami is also and advisor at Quaisr (QUantifying Uncertainty using Artificial Intelligence to Simulate Reality), an imperial College London startup from Professor Omar Matar.The startup offers a go-to-cloud integration service for powering digital twins.
Another example is a start-up in the field of generative design. Imperial College London startup ToffeAM has created an automated design software that aids the optimum design of engineering components in additive manufacturing.
Imagine a scenario where a motorcycle manufacturer equips their bikes with sensors connected to digital twins. When a part breaks on a bike, the digital twin can inform a generatively designed replacement. The tailor-made replacement could be 3D printed on the same day.
Unsurprisingly, the application of this technology has been of significant interest to industry partners such as Siemens, General Electric (GE) and Formula 1 teams.
The future of digital twins
As technologies such as quantum computing and the internet of everything (IoE) are coming of age, we can envision ever more precise digital twins. These can provide precise virtual replicas of literally anything from human bodies, entire cities, all the way up to the Earth itself.
Digital twins of the human body could be built from the DNA upwards. They could provide key information for predictive and precision medicine. Doctors could develop a digital twin of you that updates with real time data, delivered by sensors such as those in wearables or smart watches, your car or home. This virtual twin would signal and predict illnesses, enabling early diagnosis and timely and effective treatment.
Digital twins of whole cities could model our social networks, our food systems, our transportation systems. Even the air quality distribution from area to area. Not forgetting all the various primary, secondary, and tertiary complex socio-technical interactions that occur. Such as the ones among the structure of the city and its living aspects: humans, animals, plants and trees. This paves the way for reducing many of our life processes to statistical processes, that can be predicted and optimised by using digital tools.
As with many areas of modern progress, data becomes power.
Applications on the immediate industrial horizon
The fields for possible application of digital twins in the near future are vast:
- Large construction projects Buildings, bridges, and other engineered structures.
- Mechanically complex components Jet turbines, cars, and aeroplanes. Digital twins could help improve efficiency with complex machines and huge engines.
- Power grids Digital twins could help optimise mechanisms for generating power and transmitting it by modelling energy distribution grids and demand patterns.
- Manufacturing Digital twins could help streamline process efficiency, especially in industrial environments with co-functioning machine systems.
- Product design Digital twins could be virtual prototypes during the design phase and be adjusted to test different simulations or designs before building a solid prototype, reducing the number of iterations required to get to the final product
- Supply chain management logistics today could rely on digital twins to track and analyse key performance indicators, such as packaging performance, fleet management, and route efficiency.
- Healthcare Digital twins of organs in Augmented Reality settings could provide crucial information for surgeons or doctors in remote monitoring, providing possibilities for predictions and early intervention
Interested in knowing more?
Interested in this future development and what it means to you and your organisation?
Check out ” Convergence“, one of our recently released scenarios for 2041 in collaboration with Imperial College London academics
Imperial Tech Foresight is foresight backed with the scientific community of Imperial College London. Get in touch to learn more about the possibilities, challenges, and opportunities ahead with such emerging technologies.