Digital twins could save cities – and maybe your life
Digital twins are a 3D replica of an object, a process, or even a person, and have the potential to change both our physical environment and the way we live. The technology can help transform cities, give robots sight, and maybe even offer to solve the greatest of human questions: how to live longer.
To probe what possibilities the technology holds in the next decade, we spoke to Burkhard Boeckem, the Chief Technology Officer for Hexagon AB, a firm which develops the sensors, software, and platforms that power digital twins and unlock their insights.
Lots of cities have twins – usually a town of similar size in another country that allows citizens and officials to experience another culture and learn new ways of doing things. But far fewer have a doppelgänger – an exact replica of themselves – in the digital sphere.
So-called digital twins are being built not just of cities, but across a range of industries and sectors to help answer questions about supply chains, improve efficiencies, and allow manufacturers to make things better and more quickly.
The technology is possible as the physical and digital worlds become ever more connected via sensors. Combining this with improved computing power, better 3D modelling tech, and smart algorithms means the data being fed into the digital twin can offer insights about how to do things better in the real world.
Digital twins are the ideal sandbox; they let you simulate effects and changes, then derive their impact in the digital world before you deploy them into the real world.
Hexagon has worked with the Austrian city of Klagenfurt to help create its digital twin.
“They used our AI-enabled technologies to build a digital twin of the old city to find the best places to put solar panels, and in three years, they executed on this. They also thought about where the best place was for putting new green spaces and have increased those by 4%,” explained Boeckem.
In future, he thinks more and more cities will build digital twins, because the changing climate will force them to rethink their urban spaces.
“What if you don’t do it? What if there were a single mistake that you could avoid if you had seen it in the digital world? For example, we live in times of climate change, where looking at weather scenarios helps you understand and plan for future climate impacts. Digital twins are the ideal sandbox for this; they let you simulate effects and changes, then derive their impact in the digital world before you deploy them into the real world.”
Digital stadiums
Building digital replicas of our living spaces could fundamentally alter the way cities are built in the future, he believes.
The design parameters will become different, and cities will be much more friendly to live in. “I think we will see the style of buildings change, whether residential houses, offices, or industrial facilities, they will be built much more sustainably.”
And it is not just him who thinks so. Policy-makers are increasingly waking up to the benefits of digital twins. The European Commission has set aside 1.2 billion euros for digital twin projects that can improve the management and sustainability of the urban environment, while the US Department of Transport has used digital twins to predict the structural integrity of bridges, tunnels and highways.
The construction industry is also turning to digital twins to create models that are changing our built environment, with much more building happening in factories rather than construction sites. Prefabricated parts are already common for home building and will start being used for bigger projects, such as new stadiums or train stations, thinks Boeckem.
Hexagon is currently working across more than 27 industries, including automotive, aerospace, and mining.
“A mine is an ever-changing ecosystem which requires a lot of equipment – trucks, excavators, and conveyor belts to be built. And as you take the material out of the earth, it changes. By using a digital twin, you can mine for 30 or 50 years or even longer,” he said.
Meanwhile, in manufacturing, digital twins are giving industries a much better picture of what they are building.
“If you have a virtual replica of your production line, you can optimise the processes and manage the whole supply chain. You can simulate in manufacturing and design the complete product life cycle, and the digital twin can help reduce carbon footprints or increase efficiency and productivity,” said Boeckem.
For example, Jaguar Land Rover needed to renovate its older manufacturing sites in the UK and Slovakia to meet modern safety standards, fit new equipment, and minimise downtime. Hexagon helped them create highly detailed digital twins of each facility, accurate down to the millimetre, providing an exact replica of the existing buildings so that they could plan changes virtually and spot problems early. The result? Fewer surprises during renovations, smoother installations, and faster project timelines.
Digital twins are also taking product design to new heights. They are already helping to design “the perfect car” and in future will allow manufacturers to be even more experimental and bold. Jetpack, anyone? Or how about a flying car?
“Absolutely possible, because you don’t need to build prototypes. It means you can iterate faster and have many more scenarios. The second part will be down to adoption in society, but I’m quite stunned of late by how fast things can be adopted,” said Boeckem.
How will I age? Ask my digital twin
Some futurists believe that, in the next decade, humans may even have their own digital twin: a replica of ourselves which we can update with all the data we have from social media, on our smartwatches, and even the information in our work calendars. Our digital doppelgänger could take our place, attending meetings on our behalf or even providing our doctors with our latest health data.
Boeckem thinks this scenario is a little far-fetched, but says there are already plenty of examples of digital twins being used in the medical world.
“Dermatologists use AI and a copy of your face to get a prognosis of the likelihood that you might get skin cancer. And in future, we could ask the digital twin about ‘how will this person age?’ Longevity is a big trend that will also be a touchpoint for digital twins,” he said.
We see spatial intelligence as very crucial going forward, that combination of digital twins, AI, and the perception of 3D spaces. It gives you a new form to interact and understand physical spaces.
Metaverse revival?
Some have already achieved a kind of immortality via the avatars we use on gaming platforms or social media. While avatars and digital twins are not the same thing, both conjure up images of the metaverse, a much-hyped concept just a few years ago. Meta (which believed in the metaverse so much it changed its name) told us we would all be living, working and playing in the metaverse. But since ChatGPT burst onto the scene in November 2022, it has taken something of a backseat to AI.
Boeckem thinks digital twins can revive the metaverse, but its reincarnation might not happen in the way envisaged by Mark Zuckerberg and others.
“For me there are two definitions of the metaverse: the consumer version which is more about entertainment and retail, where you might have an avatar that can meet your friends in a virtual bar; and then you have the industrial metaverse, which is a combination of digital twins, simulations and AI, where you can test what-if scenarios and trial processes before you deploy them. And I think we are already in that metaverse, whether we want it or not.”
What he is most excited about for the next decade, though, is technology that will give those digital creations awareness of the real world: so-called “spatial intelligence.”
“We see spatial intelligence as very crucial going forward, that combination of digital twins, AI, and the perception of 3D spaces. It gives you a new form to interact and understand physical spaces,” he said.
“For example, in an automotive plant, you would be able to query a spatial model and tell it that you want six production lines and the output should be this many cars. And then you ask your AI to give you a layout of this. The more information you feed into the system, the more iterations and layout proposals you will get. AI and digital twins will mean you can build better and faster and with less waste, and that has huge potential. Then we will see quite drastic changes.”
All-seeing robots
Spatial intelligence would also speed up the move to human-level AI.
Demis Hassabis, the CEO of Google DeepMind, has predicted Artificial General Intelligence by 2030, but for some, this will only come with the integration of spatial intelligence.
“In 2022, we had the big breakthrough with language, and then came visuals and videos, and then multi-modal models. I think adding spatial intelligence is the logical next step,” said Boeckem.
And this will have a powerful impact on another key area of technology: robotics.
“I think that in the future, and not so far ahead, you will need a digital twin for training a robot. We will make the digital twin of the environment where the robot will be deployed and then use techniques, such as reinforcement learning, to train it. The robot has never seen the actual physical world where it will be deployed, but because it has been trained using the digital twin, it will know directions. And, in cases where the physical world changes around them, these changes can be transmitted to the robot so it can adapt in real time.”
He predicts that such all-seeing robots could be built at scale in the next decade.
Niantic, the company behind the augmented reality game Pokémon Go, thinks spatial intelligence will become the world’s next search engine. Late last year, it announced that it was building a geospatial AI model based on player data.
Boeckem agrees that, in the next decade, demand for 3D data will increase “whether that is of your sneaker, your street, or the wider environment”.
“The big question is whether the training data is as easy to find as it was for the large language models, because the world is not completely machine-readable yet. It seems likely that the next ten years will see us building a repository of digitally mapped physical spaces.”
Discover how TFD and Hexagon are working together to bring these ideas to life: https://www.wearetfd.com/what-we-do/hexagon-geosystems
