Demystifying Digital Twins
Opportunities and challenges
The TDL webinar that was broadcast on 14 May was intriguingly titled “demystifying digital twins.” Digital twins are virtual replicas of physical objects, processes, systems or even entire environments that use real-time data from sensors (IoT), AI, simulations and modelling to mirror, predict and optimise the behaviour of their real-world counterparts.
Originally emerging as a way to formalise complex structures in digital form, digital twins went from being useful to having a prominent role over the course of the last five years. Their foremost role is in manufacturing and supply chain where models of manufacturing assets and processes are now commonly used to manage manufacturing facilities and processes and to represent supply chains, especially when digital and physical products and operations need to be evaluated together. More recently, digital twins branched out into a large number of additional fields, from defence to healthcare.
Digital twins present opportunities for optimisation and innovation, risk analysis and quick responses to issues, but at the same time, the approach has numerous challenges, including costs, accuracy, uncertain effect on ROI and organisational resistance.
The issues the eminent speakers were asked to address included what the present and future of digital twins is, whether they will move from promising but experimental techniques to approaches that are broadly applicable and what areas will they thrive in.
Summary
This panel discussed demystifying digital twins, focusing on the evolution of digital twins from their origins in space missions to their current applications in manufacturing, aviation and various other industries. They explored the three stages of digital twin development, from static 3D models to sensor-connected systems and physics-informed decision engines. The participants identified key drivers of growth including sensor technology, AI advancements and cloud-based deployment, while highlighting challenges such as data integration, cybersecurity and the need for standardised interfaces. The panellists expressed optimism about the future of digital twins, particularly in their potential to democratise knowledge, enable startup innovation and help companies avoid costly downtime, though they acknowledged challenges around accuracy, data collection and system integration.
Background
The topics to be addressed in this focus on digital twins included the status of the technology, its prospects, its challenges and the main areas of application, both current and in the future.
Among the general issues arising and getting to the heart of the matter, the first question is whether real time synchronization - data flow between digital and physical assets - is used in many areas outside of the labs today, as well as under which circumstances digital twins are the best for complex simulations. Looking ahead, it is important to establish what will yield the most promising outcomes in terms of types of digital twins are - a process, a system, etc, areas of application and research directions as well as understanding what the major challenges are.
Main drivers and challenges
The market projections for the technology point to significant growth but at this relatively early stage it’s not obvious how realistic this positive prognosis is without getting a better take on what the main operational and technological drivers for the adoption of digital twins are, such as increased efficiency, real time monitoring, simulation of complex processes, availability of AI tools, etc as well as what the economic drivers are, such as the low cost of IoT and cloud aggregation, AI, etc. And looking at the barriers to adoption, among the key obstacles to deployment are, for example, cost, complexity, security issues and regulatory challenges.
Major technology innovation in digital twins
At present there are only a few areas, for example, manufacturing and medical care, that have produced or are producing the most interesting technology innovation in connection with digital twins and it’s of interest to know whether AI has been the source of most innovation from data processing and signals to faster optimisation. Looking to the future, there may be opportunities for digital twinning to be used more in complex designs, such as for semiconductors. But it’s not apparent whether digital twins are a relevant approach for consumer services, like virtual reality environments. Scientists and practitioners working in this field are still grappling with what the most promising research directions are and what operational innovations might be connected with digital twins.
Longer term prospects and limitations.
Having established that this technology is still, from an operational perspective, in its infancy, there are open questions about its longer term prospects and whether it will bring significant optimisation to mission critical functionality or remain supplemental. Finally, the panellists were asked what futuristic applications of digital twins they envisioned and what needs to be done to ensure the market and research projections on digital twins become reality.
More questions than answers, a good place to start a discussion!
Speakers
The distinguished group of experts were:
· Ashis Khan, CEO/Founder, Intelligent Shopfloor, a Silicon Valley startup
· John Oakley, Professor of Practice in Computer & Information Science & Engineering (CISE), University of Florida
· JV Rajendran, Associate Professor and ASCEND Fellow at Texas A&M University
The session was moderated by TDL strategic adviser, Claire Vishik
Digital Twins Evolution and Applications
The meeting focused on discussing digital twins, their evolution and current applications, starting with an explanation as to how digital twins have evolved from static 3D models to connected systems that can predict and prescribe actions in the physical world. The challenges inherent in implementing digital twins were highlighted, including cybersecurity risks, data integration issues and difficulties in scaling across different plants and assets. Digital twins were defined as virtual models informed by the real world, emphasising the importance of bi-directional feedback between virtual and physical systems. The discussion touched on the need for clearer definitions of digital twins and their applications across different industries.
Digital Twins Industry Prospects
The panellists discussed the current state and future prospects of digital twins across various industries. The potential for digital twins to help with hardware security and multi-physics effects in chip design was emphasised, highlighting the role of sensors, AI and cloud services in driving growth. One of the fascinating areas provided by AI and transforming digital twins is the democratisation of knowledge. As always in the complex systems, integration is an issue and the challenges of connecting different digital twin systems are the main obstacle to deployment, along with cost and operational problems. The panellists agreed that while there are challenges, including standardisation and data collection, digital twins have significant potential for growth in manufacturing, transportation and other fields. They also discussed the importance of cybersecurity and the need for tiered models to make digital twins more accessible to smaller players. Overall, the panellists expressed optimism about the future of digital twins as a key technology in industry operations.
Watch the full recording of the webinar on our YouTube channel.