A smart city manages urban operations using data from infrastructure and buildings to inform decisions. During construction its foundation is laid through consistent digital models, long before any sensor or monitoring system is installed.
Why does a data-driven city start with the model?
A monitorable city system needs one thing before any sensor is installed: consistent asset identity. Without uniform naming and data structure, data from different sources cannot be brought together, and every system ends up an island that cannot be compared with any other.
The BIM model supplies that structure from the design stage. Every component carries an identifier, attributes and spatial context, so information gathered during construction can be used in operations without being rebuilt.
The sequence cannot be reversed. Installing sensors on assets whose identity is inconsistent produces a data stream that attaches to no component, and reconciling it later means mapping thousands of data points manually, one at a time.
What changes for contractors?
For contractors used to conventional projects, the change touches deliverables, site record-keeping, and the definition of finished work. Four things are felt most directly.
| Aspect | Conventional project | Data-driven project |
| Final deliverable | As-built drawings | As-built drawings and structured asset data |
| Component naming | Free per discipline | Bound to a single project standard |
| Handover | Printed or PDF documents | Data a management system can read |
| Quality checking | Focused on physical works | Physical works and completeness of data attributes |
The biggest change is not software but site record-keeping habits. Data not captured during construction is nearly impossible to reconstruct afterwards — installed equipment serial numbers, installation dates, and deviations from the drawings are known only to whoever was on site that day.
Practically, quality checking gains a dimension. A work package is no longer complete because it is physically installed; it is complete when its data attributes are populated and have reached the shared data environment.
What readiness must be built first?
Four items are prerequisites, and their order cannot be swapped. The first three must be settled before modelling begins, not agreed while it runs.
- Project information standards agreed before modelling starts, including naming rules and data structure.
- A shared data environment where all parties work, with a functioning status and approval flow.
- Agreement on which attributes attach to each asset type — and who populates them, at which stage.
- Clarity on who maintains the data after handover, since data left un-updated loses its value within months.
The third is most often agreed too late. Defining mandatory attributes after modelling is under way means returning to thousands of existing elements to complete them, and that work is almost always rushed against handover with quality to match.
What does the digital twin do after construction?
Once the building operates, an accurate model can be connected to building system data so asset condition is readable without a site visit. That is where the investment in data discipline during construction starts to return.
The benefit is felt most in facilities with dense mechanical and electrical equipment, where maintenance scheduling drives operating cost. At estate scale, being able to compare performance between buildings on the same basis becomes an additional benefit — one unavailable when every building hands over data in a different structure.
What goes wrong in the early stages?
Four patterns recur in nearly every data-driven city initiative, and all four stem from a reversed sequence.
- Installing sensors before asset data is clean, so readings cannot be tied to any component.
- Producing a model that looks impressive but carries few attributes, making it useless for operations.
- Deferring naming agreements until construction is running, by which point changing them is expensive.
- Treating data handover as end-of-project administration rather than a deliverable planned from the outset.
The second pattern deceives most, because the result looks convincing in a presentation. A visually strong model with no equipment attributes answers no operational question, and the shortfall only surfaces once the facility team tries to use it.
Frequently Asked Questions
Is smart city only for new cities?
No. The same approach applies to an estate or a group of buildings under one operator, at far smaller scale. What matters is not size but whether asset data uses a shared structure so it can be compared and combined.
Smart city vs smart building?
A smart building concerns one building; a smart city brings together many assets and infrastructure under broader management. The data principles are identical — consistent asset identity and uniform attributes. The difference is scope and the number of parties who must agree on the same standard.
Is BIM mandatory for government projects?
PUPR Regulation 22/PRT/M/2018 mandates it for state buildings classified as non-simple, above 2,000 square metres and more than two storeys. All three conditions apply together, so not every government project falls inside it. Data-driven programmes such as new city districts typically demand more than that minimum in any case.
Should contractors prepare now?
Yes, and preparation does not wait for a data-driven project to appear. The habit of modelling with consistent naming and complete attributes can be practised on projects running today, and that is the baseline capability assessed when such projects open.
Sources
- Permen PUPR No. 22/PRT/M/2018 tentang Pembangunan Bangunan Gedung Negara — JDIH Kementerian PUPR — https://jdih.pu.go.id/detail-dokumen/2594/1
- ISO 19650-3:2020 — Operational phase of the assets — https://www.iso.org/standard/75109.html
- ISO — Building information modelling (BIM), seri ISO 19650 — https://www.iso.org/sectors/building-construction/building-information-modelling
Get an Initial Consultation with BIMAGE Indonesia
BIMAGE Indonesia supports information standards, modelling and asset data handover for projects that demand operational data readiness, including digital twin delivery on the BIMAGE 360 platform. If you are building capability for data-driven government work, naming and attribute readiness is the cheapest place to start — bring your current internal standard to an initial consultation.