Why CAFM imports fail, and which standards prevent it
Most failed CAFM data handovers fail not because of the volume of data but because of four small things: ambiguous room numbers, drawing and list that spell differently, attributes that do not fit the system’s data model, and areas that appear only as text in the drawing rather than as a property of the room polygon.
There is no software against this, but there are standards. Anyone who knows them can recognise a dependable data supplier by three questions.
Question 1: in what structure does the data arrive?
The basis is DIN 277 for the areas and DIN 276 (cost group 400) for the technical services. Above that lies the question of format: IFC to ISO 16739 as a vendor-neutral model format, and for the FM context CAFM Connect, the standard maintained by the CAFM RING, which builds on IFC and structures room and plant data so that CAFM systems can read it without special arrangements. On internationally oriented projects, COBie is added as a tabular form of handover.
Anyone who only says “DWG and Excel” has not answered the question. What matters is how rooms, areas and attributes are linked within them.
Question 2: how does the supplier know how much information is enough?
That is what the level of information need is for, standardised in DIN EN ISO 7817-1: the information requirement is defined from the intended use, geometrically as well as alphanumerically, instead of ordering “as detailed as possible” across the board. In practice that means agreeing the level of development per trade and deriving the scope of attributes from the data model of the target system – both before the survey.
The German BIM guideline series VDI 2552 supplies the accompanying vocabulary. For choosing the system side, in turn, GEFMA 444 is the market anchor: it certifies CAFM software products against catalogues of criteria, among them BIM data processing. Anyone supplying data should know what has been certified on the receiving side.
Question 3: how is it proved that the import works?
Not with a promise but with a test. The dependable sequence: sample drawing and test import into the target system on a pilot building, final agreement on layers, stamps and attributes, and only then the series.
That is how we work on every portfolio project, and it is the step that tests the four small things from the start out of the project while they are still a correction and not a remediation case across forty buildings.
Further reading
- CAFM data basis – the service behind this article
- Room book – the alphanumeric half, linked by the room number
- BIM as-built models – where the level of development is agreed
- Area calculation – areas as a property of the polygon, not as text in the drawing
- Prices – what makes a data delivery comparable
Frequently asked questions
Does the CAFM system have to be chosen before the survey?
No. A vendor-neutral structure (IFC, DWG, Excel to DIN 277) keeps the data independent of any system. Once the system is settled, the data is additionally prepared against its data model and the import is tested.
Is IFC alone enough?
For the geometry yes, for operation rarely. FM systems need the alphanumeric half, cleanly linked by the room number. That is why the room book and the drawing belong to one source.
What most often makes an import fail?
The room number. In stock that has grown over time, several numbering systems exist side by side, and which one applies is decided by the operator, not by the data supplier. That decision belongs before the survey.