Visualising the Twin

The same twin can be shown through very different windows, and it helps to hold them as a set rather than assuming 3D is the only option. Each window is a trade between reach, richness, cost and the kind of question it answers well.

Real-time 3D in a game engine is the immersive, photoreal, interactive view - you can walk or fly through the city, change the time of day, watch a simulation play out in space. It is unmatched for engagement, design review and making a scenario feel real to a committee or the public. Its costs are real: it is heavy to build, demanding on hardware, and expensive to keep current as the city changes. A web viewer, using WebGL and streaming formats like 3D Tiles, trades some richness for enormous reach: it runs in an ordinary browser with no install, so anyone with a link can see the twin. It is the right window when sharing and public access matter more than cinematic detail, bounded by what a browser and a phone can handle.

VR and AR push toward presence. Virtual reality puts you bodily inside the model, which can be powerful for experiencing a space before it exists; augmented reality overlays the twin onto the real site through a phone or headset, so you can stand on a street and see the proposed building, or see underground pipes beneath your feet. Both are compelling for specific tasks - experiential design, on-site inspection - but reach a narrow audience and carry hardware and comfort costs, so they are tools for particular moments, not the everyday view.

2D dashboards, charts and maps are the window people forget because they are not glamorous - yet they are how most actual decisions get made. A ranked list of alerts, a trend line, a choropleth map of air quality reads faster and more honestly than any fly-through, and it is cheap to build and maintain. The mistake is to treat the 2D dashboard as a lesser thing, or conversely to treat the 3D as the whole twin. The discipline is to match the window to the question: immersive 3D to make people feel a proposal, a web viewer to share it widely, VR/AR for presence and on-site work, and 2D for the monitoring and analysis where speed and clarity beat beauty. Most mature twins use several windows at once, each doing what it is best at.

Here is the tension at the heart of twin visualisation: the most beautiful view and the most useful view are often different, and the pull toward beauty is strong because beauty is what impresses funders, committees and the press. A photoreal fly-through of a glowing city looks like the future and feels like progress. But if it does not help anyone answer a question or make a decision, it is, however gorgeous, decoration - and decoration that cost a fortune to build and will be obsolete the moment the city changes.

The opposite failure is just as real. A blunt, ugly chart or a plain map can be so austere, or so badly designed, that the people who need it ignore it, and an insight nobody looks at changes nothing. So the goal is not to reject beauty in favour of plainness; it is to reach the narrow, valuable band where a view is both clear and useful - legible enough that people engage with it, honest enough that it answers a real question without misleading. A good traffic dashboard and a good design-review fly-through both live in that band; a gratuitous cinematic showpiece and an unreadable wall of numbers both fail it, from opposite sides.

There is a quieter danger in beauty specifically: it can hide uncertainty and lend false authority. A crisp, confident, photoreal image of a simulated flood or a proposed skyline looks definitive, but the data behind it may be incomplete and the simulation may carry large error bars. The more polished the picture, the more it invites people to forget that a twin is a purposeful simplification fed by imperfect data - exactly the false confidence this course keeps warning against. Honest visualisation does the opposite: it shows uncertainty, marks what is measured versus modelled, and resists making a guess look like a fact.

The practical test, every time, is a single question asked before the view is built: what decision is this supposed to help someone make, and who is that someone? If you cannot answer, you are not designing a visualisation, you are designing a demo. Answer it, and the right window, the right level of polish, and the right honesty about uncertainty usually follow. Beauty is welcome, but only in service of insight - never as a substitute for it.

Step back and the real subject of this lesson comes into view: a twin's visualisation is an act of communication, and every communication has a sender, an audience, a message and a purpose. Treating a view as mere output - 'show the data' - misses that every choice in it is an argument about what matters. What you put at the centre, what you colour red, what you leave out, what scale you use: these are editorial decisions that shape what the viewer concludes. A designer who understands this takes responsibility for the argument the view is making, rather than pretending the picture is neutral.

That responsibility runs in two directions. The first is toward clarity for the audience: a view built for a traffic operator who has thirty seconds and needs to know what to do must look nothing like a view built for a planning committee weighing a decade-long decision, or for a citizen trying to understand a proposal near their home. The same twin, honestly, should produce different views for different people - a theme lesson 5.3 develops for dashboards. Designing one dazzling view and expecting everyone to use it is a failure of communication dressed as efficiency.

The second responsibility is toward honesty. Because visualisation is persuasion, it can mislead - through a colour scale that exaggerates, a 3D perspective that hides what is inconvenient, a confident render that buries uncertainty, or a carefully framed fly-through that shows the development from its best angle and never from the blocked view of the existing neighbours. In a civic context, where a twin's images shape real decisions about real people's lives, this is not a stylistic matter; it is an ethical one. The view can be used to genuinely inform a public debate, or to manufacture consent for a decision already made. Knowing the difference, and refusing to dress persuasion as fact, is part of being a trustworthy designer in this field.

So the craft of visualising a twin is not primarily a graphics skill; it is a communication skill with a graphics component. Know your audience and their decision, choose the window that serves it, reach for the band that is both clear and useful, show uncertainty rather than hiding it, and remember that the picture is always making an argument - so make an honest one. The most sophisticated rendering pipeline in the world cannot rescue a view that has forgotten whom it is for and what it is trying to say.

How does this translate into practice for a designer working with or contributing to a twin? A few habits separate visualisation that earns its keep from visualisation that merely impresses.

Start from the decision and the audience, never from the tool. Before choosing a game engine or a dashboard library, write down who will look at this and what they are trying to decide. That single discipline prevents most waste, because it rules out the cinematic showpiece that answers no question and the data dump nobody can read. Then choose the window to fit - immersive 3D where feeling the space matters, a web viewer where reach matters, VR/AR for presence and on-site work, 2D for speed and analysis - and be willing to use several windows for several audiences rather than forcing one view to do everything.

Design for legibility and honesty together. Use colour and emphasis to guide attention to what matters, but choose scales and framings that do not exaggerate or mislead; label what is measured versus modelled; and find a way to show uncertainty rather than letting a confident image imply a certainty the data does not support. Keep the view maintainable: a twin changes, and a beautiful visualisation that cannot be kept current becomes a lie within a year, so favour views you can actually update as the city and the data move. And treat performance and access as design constraints, not afterthoughts - a view that only runs on a powerful workstation excludes most of the people a civic twin should serve, which is both a usability failure and, often, an equity one.

Finally, keep the whole thing in proportion. Visualisation is the window onto the twin, not the twin itself; the value lives in the model, the data and the decisions, and a gorgeous view on top of weak data or absent governance is twin-washing with good production values. The designer's job is to make the twin's genuine insight visible, usable and honest to the people who need it - and to resist, every time, the seduction of making it merely beautiful. Remember too that any specific capability, resolution or accuracy a visualisation seems to promise is illustrative and context-dependent; the binding facts about the city still come from the official data and the accountable authorities, not from how convincing the picture looks.