Continuous product improvements drove the recovery through 2023 and 2024, and in January 2025 we opened embedded AI to all new customers. From that point, the decline has accelerated even further, with the downward trend continuing into 2026.
The numbers tell a clear story. Across 48,509 customer observations and 6.9 million tracked failures between 2020 and 2026, median MTTR has dropped 42 percent, and three in four customers who were with us across 2024 and 2025 saw their repair times go down. That's not a marginal gain; it's a structural shift in how fast maintenance teams can diagnose and resolve failures.
Why the trend is accelerating
Embedded AI works because it removes friction at the exact moments technicians lose time: writing reports, interpreting photos, and translating findings into next steps.
AI enabled failure reports suggest relevant symptoms and observations based on the selected asset, so technicians pick from smart suggestions instead of typing everything from scratch, resulting in richer, more consistent data.
Automatic image descriptions turn uploaded photos into searchable, structured information, giving teams instant visual context without extra effort.
Assisted troubleshooting uses vector-based search to surface similar past failures, grouping them by root cause so technicians can act on proven fixes instead of starting from zero.
AI generated work instructions and inspection lines turn short planner notes into structured, step by step guidance, cutting preparation time and standardizing documentation across sites.
Each of these features attacks a different stage of the repair cycle, which is exactly why the effect compounds. Better initial data means faster diagnosis, faster diagnosis means less downtime, and less downtime is what shows up in the MTTR chart quarter after quarter.
The takeaway for maintenance leaders
If your team is still facing the same MTTR plateau from the post COVID years, the data suggests the ceiling has moved. Organizations running (embedded) AI aren't just recovering to pre 2022 performance; they're setting new MTTR lows.
Customer stories back this up. The Technical Services team at Berkvens Doorsystems reports saving 30 to 60 minutes per team lead every day now that information is automatically summarized, with AI generated insights matching their own analysis more than 95 percent of the time.
Infographic
We have designed an informative infographic. You can download it for free.
More information
Want to see how Embedded AI could move your own MTTR numbers? Explore the full feature set or request a demo to see it working on your assets.
