Knowledge base for AI ·
From building data to shared knowledge for AI
Building data becomes more valuable when professional assessments accompany it. A shared knowledge base can help the next colleague or AI tool build on the work.
Smart Campus NMBU links building data to the same rooms and technical systems. SC Insight uses this foundation for analysis and decision support. A possible next step is to preserve the knowledge gained when professionals investigate the data.
Keep the reasoning alongside the data
Ventilation running at night may appear to be an anomaly. Yet laboratories or other processes may require it. The analysis therefore needs professional judgement that considers local needs and data quality.
If an investigation establishes why the system needs to run, that reasoning is useful next time too. Link it to the relevant system, period and source, recording who assessed it and what remains uncertain. Hypotheses, confirmed experience and current requirements need clearly distinct statuses and must be open to revision.
Help the next person continue the work
AI can suggest summaries for a professional to correct and confirm. Preserving findings and open questions outside individual conversations can make them available to other colleagues and tools. Sources, access permissions and responsibility must stay attached, and the workflow must ensure that knowledge is actually saved and can be retrieved.
The diagram from Skogli digital, translated into English, shows a proposed development: relevant knowledge gained through the work feeds back into a shared foundation.
This is a direction for further work, not current SC Insight functionality. Start with one investigation: can the next colleague find the reasoning, understand what has been done and continue where the work left off?