NASA & rockets, very good case where you would throw a couple 100 grand on it to decrease the likelihood of failure even more

Actually, the PdM is for all of their facilities and equipments rather than the actual rockets. I think there's still applications where machine uptime and unplanned failure costs would make this a valid solution. My problem is that most maintenance software are basically just a maintenance based MES with little actual PM/PdM etc built into it.

Actually, I saw this in my research. As @Jermuk said, the vibration crossed a certain limit and generated an alarm. It looks like PdM. My approach is to determine the date and point of the failure based on anomalies. For this, a lot of information such as current sensors, cutting tool information, vibration information needs to be processed. But the need for this information is machine-based. So something is needed that can easily aggregate data and communicate with PdM. There MC comes to my aid 🙂

This is what some of the PdM-systems does to some extent.. this i also what i would like to accomplish with the UMH in the future.. both PdM for the machines but also anomaly detection where we can use historic machine data to see for example if the quality is reduced or if the heating time takes longer for a specific product, then we can find errors in the machine long before a crisis..