One year into running autonomous maintenance in production, 42% of work orders are now guided end-to-end by AI, with a single reviewer holding oversight across 30,000+ triaged faults. For CMMS-governed facilities management (FM) operations, this is the proof point that AI in FM can move past pilot, not as automation bolted onto a ticketing tool, but as a governance layer that makes the first call on every fault and shows its reasoning. 

‍

The clearest signal is in the escalation rate. Early on, nearly every fault required human triage before any action. Today, only 1 in 7–8 faults needs a reviewer in the loop, the rest are diagnosed, prioritised, and routed autonomously. The cases that still escalate are the ones that should: edge conditions, ambiguous diagnostics, and calls where the model defers by design rather than by failure. 

‍

The practical shift for FM teams is in the supervisor's role. A year ago, the job was triaging every alert as it came in. Today, the same reviewer governs the AI's calls across an entire portfolio — auditing the reasoning, intervening on the 1-in-8 that genuinely need judgment, and trusting the system on the rest. That is what scaling FM operations without scaling headcount actually looks like in practice. 

‍

For the full framework behind these numbers, see: What Is Autonomous Maintenance? → 
‍

FAQ

How is autonomous maintenance different from CMMS automation?

A: CMMS automation follows rules someone has to write, if this happens, create that ticket. Autonomous maintenance lets AI make the first call on each fault and show its reasoning, while a human reviewer governs the calls instead of writing rules for each one.