AI is enabled by location data

AI in property management is often discussed as a standalone solution. That is the wrong starting point.

AI becomes valuable when it has the right context. And context starts with location.

A work order without a location link gives AI about as little to work with as it gives a property technician: a free text string that has to be interpreted manually. Knowing where the fault is, what it concerns and what has happened before, that is AI's raw material. Without it, it cannot enrich, prioritise or suggest.

With location data, the situation looks completely different. The work order is linked to a room and a floor. That room is linked to a ventilation unit. The unit has a service history, an expected lifespan and a number of previous work orders. AI can now do something meaningful: combine sensor readings with history, identify that the high CO₂ level combined with low air flows points to a unit problem rather than tenant behaviour and suggest the next step.

What looks like AI magic is really integration and location logic. AI is most needed when information is unstructured or contradictory, when several data sources need to be weighed together and when patterns need to be found in large volumes of historical work orders. The structured groundwork, location, asset, history, is the prerequisite.

A reasonable target is a digital team of specialised agents: one that interprets the report in text, image or speech, one that enriches the work order with information from existing systems, one that assesses and proposes a course of action and one that follows up outcomes and identifies patterns over time.

A particularly underused opportunity is reporting back. The technician can dictate what was done and add a photo, and AI structures the information into cause, action, component, time and cost. Every closed work order leaves behind useful knowledge, instead of the organisation starting from scratch the next time the same unit fails.

The organisations that build that foundation now will have far more opportunities to use AI as the technology develops. Those that wait will have more raw data, but not more knowledge.

Vyer has, together with Vasakronan, Jernhusen, Algorithma, Solrun and Epicenter, contributed to a report published by REDI on work request management and AI. Read about AI and much more in the full report at redi.city/rapporter/visuell-felhantering-och-ai.