Calendar schedule of when parts were replaced does not directly represent their life.

Were they actually replaced because they wore out from use or other factor - e.g. some components break down over time even if the machine is not being used.

What was their duty cycle, 24/7/52 or 8/5/47 or ....

Is their wear even through out use, or at certain stages do they see more wear e.g. at startup/shut down?

Are the machines all being used similarly e.g. a tool operating at 10% capacity may last longer than one working at 100% capacity in any critical spec.

Even if you got a massive amount of data of how long between changing parts, if the relevant usage conditions aren't taken into account, it's difficult to come up with a good life estimate. Best you could take a conservative value e.g. 5th percentile and use this as 'typical life dependent on usage conditions' or something.

You need to reduce the variables..
As stated by KENAT tool life can vary greatly depending on many..many factors (variables)..

Do you have hour meters on the machines ? If you can record hours as well as stop start cycles it will help.
It would also help if you told us what type of machines they were. Are you dealing with conveyors, loaders, milling machines, lathes, bearings, motors etc.
The type of machine greatly affects the duty cycle.
B.E.

You are judged not by what you know, but by what you can do.

What kind of machine parts? Machine tools? Or what?
If it is machine tools that would be close to impossible to determine. Why, because too many variables, like are you using the default speed parameters or did you speed them up or slow them down? What kind of loads are you putting on the machine? What kind of materials are you cutting on the machine? What coolants are you using? What are the conditions in the shop? If you want to know how long your machines will last, start calling on other shops that use the same types and brands of machines, and you will get a good idea.