CAD and engineering twins mostly freeze the geometry
When a vendor hands over a CAD or BIM export and calls it a digital twin, the data that enters the system is almost always design geometry. Files arrive as .fbx, .obj, IFC or glTF. Engineers may polish the model, cut polygon counts or split floors so the file stays under practical size limits. Once loaded, that geometry stays put. The next update arrives only when someone re-exports a revised layout after a physical change on the floor. No continuous stream of forklift coordinates, material locations or PLC states feeds the model. The twin shows the planned shape of the plant; it does not show where assets and people sit at this moment.
Point clouds, if used at all, first have to be meshed before they can appear. The architecture never claims to ingest live operational feeds. Its strength sits in the fidelity of the static model, not in the freshness of floor data.
IoT dashboard twins keep sensor values current but leave space out
A second group of platforms collects live numbers from sensors, PLCs, SCADA systems and APIs that already sit inside MES, ERP or WMS. Temperature, current, cycle counts, downtime codes and KPI values refresh as soon as the source systems publish them. The architecture therefore updates in real time. The display, however, remains a dashboard or a set of time-series charts. The position of a forklift that just finished a move, or the exact bay where a micro-stoppage began, never appears inside a spatial model. Real-time data arrives; spatial context does not.
These systems excel at keeping numbers current. They do not place those numbers back into the physical layout of the plant. A supervisor still has to translate a rising temperature or an open work order into a physical location by memory or by walking the floor.
Spatially fused RTLS twins keep both position and process data live
The third architecture starts with the same kind of 3D geometry yet continuously ingests real-time location feeds. Platforms such as Twinzo take positions generated by BLE, UWB, RFID or SLAM systems and place the moving icons of forklifts, people and materials directly inside the model. Position packets typically refresh every few seconds. At the same time the platform pulls production metrics, sensor readings and system status from existing MES, ERP, WMS and SCADA systems through APIs and edge protocols that include OPC UA, MQTT, Modbus TCP and Ethernet/IP. The result is a single live map in which both location and operational state stay current.
Because the location layer is independent of the process layer, the same 3D scene can show a forklift’s current coordinates next to the line status or material request that triggered its last move. Historical position data remains available for replay, so engineers can later examine routes, dwell times and spaghetti diagrams without having to reconstruct the shift from memory or from separate log files. The architecture does not replace the source systems; it simply overlays the spatial layer on the data those systems already produce.
What actually refreshes on the screen
In the CAD style the only refresh is a new model file after an engineer finishes an update. In the IoT dashboard style the numbers and charts keep moving while the background layout stays fixed. In the spatially fused RTLS style both the moving asset icons and the overlaid process values update together.
The difference shows up the moment a supervisor needs to know not only that a stoppage occurred, but where the nearest available forklift is and which material request is still open. Only the third architecture keeps both pieces of information current inside the same view.
Matching the data feed to the decision that needs to be made
If the open question is still about layout options before steel is cut, a CAD or simulation model supplies the right data set. If the daily need is simply to know whether a machine is running or a temperature is rising, an IoT dashboard already delivers those numbers. When the recurring problem is knowing where every forklift sits right now, how material is actually flowing across the floor, and how those movements line up with line status or open work orders, the architecture that keeps both position and process data current in the same view becomes the practical choice.