A city plan often fails in the room rather than on the drawing board. A council member looks at a two-dimensional map, a contractor reads a cross-section, and a resident sees only a rendering that hides the street they actually walk. Each person leaves with a different picture of the same project. Holographic and augmented reality views of a digital twin address this gap by putting one shared model at the centre of the discussion, at a scale and height that everyone can read.
In practice, the workflow starts with the same data that drives the operational twin: surveyed parcel boundaries, building heights, road geometry, utility lines and live sensor feeds such as traffic counters or air quality stations. The model is then exported to two display modes. A holographic table or room projection suits group sessions, where several people walk around a block-sized model and point at details. A headset or tablet AR view suits site visits, where a planner stands on a pavement and sees a proposed tram stop or pocket park overlaid on the real street, with the correct sightlines and shadows for the time of day chosen.
A realistic scenario is a municipality considering a new junction layout. In a conference room, the team can show the current signal timings and queue lengths as animated columns above each lane, then switch to the proposed layout and let officials compare the two views in the same frame. On site, shopkeepers can open the AR view on a phone and see whether a planned loading bay would block their entrance. Questions that would have taken several meetings to surface often come up within the first ten minutes, because people are reacting to the same visible object.
The value depends on discipline. Every overlay should carry its source, date and confidence, so that a projected queue estimate is not mistaken for a measured count. Scales must be stated, and the holographic model should be checked against the survey so that a building does not appear taller than its permit allows. Sessions work best with a short agreed list of questions written before the viewing, and with the decisions and open items recorded against specific objects in the model rather than in loose meeting notes.
There are also limits. Holographic displays are expensive to install and are best used for a few key decisions, not for every routine review. AR overlays drift if positioning is poor, so site views need calibration against fixed survey points. And visual realism can persuade people more than the data justifies, which is why facilitators should show uncertainty bands alongside the attractive renderings.
For teams planning such sessions, the most useful starting point is modest: pick one contested decision, build the view from data already maintained by the city, and test it with a mixed group before scaling. Platforms such as TwinUp's Clavus City are one way to manage these shared models, but the technique works equally well with any reliable, well-governed twin.