- within Technology, Immigration and Strategy topic(s)
- in United States
People have been talking about digital twins for decades. Finally, the pieces are in place to turn models into value.
Vindication, at last. For more than 30 years, promotors of digital twins have been prophesying their ascendency (regular readers of Emerging Trends in Infrastructure and Transport will know we're among the frustrated prophets). Yet uptake has been disappointingly slow, for good reasons.
Only a few years ago, the cost and complexity of developing a good digital twin was sky high. It took years just to collect the data, standardize it and drop it into a data lake. The software tended to be highly-technical and required legions of programmers and data scientists to set up. Even then, a lack of interoperability between systems often meant that scope was limited. As a result, there were few truly valuable use cases. The cost and complexity generally outweighed the benefit. Those who were able to build a business case often got bogged down trying to develop the 'perfect' twin.
Over the past few years, however, many of the barriers have fallen away. Technology has matured to a point where, generally speaking, data can be scooped up from almost any source and fed into the system in realtime. Coupled with new data centers, cloud computing, 5G technologies and increased processing speeds, many of the challenges organizations faced wrangling data have largely been solved.
Over the past few years, many of the barriers have fallen away.
At the same time, the software and platforms have become much more user-friendly. Building a good digital twin no longer requires a PhD in Mathematics. Nor does using or operating one. More recently, we have seen an explosion of platforms that provide pre-built maps based on LIDAR, BIM, GIS and other existing data sources, enabling businesses and public sector leaders to roll out digital twins rapidly and cost-effectively. What is more, interoperability has vastly improved, in part thanks to greater standardization across devices, sensors and data.
Attitudes have also changed. Leaders are no longer preoccupied with creating a perfect and all-encompassing digital twin. Instead, they recognize that they can get an awful lot of value from an MVP (Minimal Viable Product). The consumerization of tech has also shifted decision-maker expectations. If I can watch my pizza travel from store to door in real-time, why can't I watch a truck leave my warehouse?
Reset 2025
Vindication should come quickly. Expect to see digital twin adoption moving both from the top down and the bottom up. On one side, a nice trifecta of lower cost, decreased complexity and increased demand for insights from executives will drive a greater level of enterprise adoption. At the same time, the democratization of digital twin technologies coupled with generative AI will allow desk jockeys to build their own library of assets. Both should be embraced.
If I can watch my pizza travel from store to door in real-time, why can't I watch a truck leave my warehouse?
Indeed, advancements in AI and machine learning will make digital twins even more predictive and autonomous, while the integration of IoT devices will further enhance real-time data collection, enabling organizations to monitor and optimize their operations with unprecedented precision.
Rapid uptake is already underway in complex, asset-heavy sectors like airports, ports and manufacturing plants. As suppliers and partners start to integrate their data into existing models, we should start to see adoption of digital twins diffuse down the supply chain.
We believe that — within the next three years — digital twins will be widely regarded as central to business decision-making. Digital twins will drive operational decisions (often autonomously) and will drive significant productivity gains through predictive maintenance. They will provide leaders with the organizational spatial awareness they need to make smart long-term decisions. They will enable better scenario planning, simulation and — as a result — capital investment planning. Digital twins will be the lens through which smart leaders view their organizations.
Those without digital twins of their core assets or operations will face questions about their control, governance and reporting.
Advancements in AI and machine learning will make digital twins even more predictive and autonomous.
Our prediction and advice
It won't be long before key stakeholders — regulators, investors, suppliers and customers, for example — will start to expect digital twins to be part of an organization's day-to-day decision-making. Those without digital twins of their core assets or operations will likely face questions about their control, governance and reporting. They should also expect to face competitive disadvantages as their peers move ahead with their models.
That being said, technology should never lead the business case. The digital twin journey should always start with a clear understanding of the challenge you need to solve. And it should be designed in cooperation with the endusers and operators to ensure it delivers on its objectives. With the end-user and business objective in mind, the choice of technologies, data sources, analytics engineers and visualization techniques can become much clearer.
While, in some cases, perfection is absolutely necessary (ask NASA), decision-makers should also take the time to decide what level of detail and certainty they actually require in order to achieve their objectives. Know what's 'good enough' and go for it.
For those of us who, for decades, have prophesied the impending rise of digital twins, the next few years will be truly vindicating.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
