Hi Roxy
Your post has been read over 60 times but without reply, so I will try and get the ball rolling.
I think you might need to expand on the nature of the question.
All too often "fatigue management" is done by counting the number of hours people work in a day/week/longer period, partly as a nod to whatever legislation is in place to comply with the EC Working Time Directive and such like.
However, there are many other things to consider.
As example, there tends to be the assumption that those working night shifts will inevitably suffer from "fatigue" but studies long ago indicate that many people adapt entirely comfortable to working different hours to the majority, and that it is CHANGES to shift pattern that are sometimes likely to be more of an issue.
So another example, you can put people on rotating shifts say 8am - 4pm one week, 4pm to Midnight the next, Midnight to 8am the third and then repeat and many workers will probably adapt quite well.
In contrast, decades ago, researchers found that many people could not cope with the so call Reverse European Shift Pattern, so instead of the shifts going forwards from e.g. 8-4 to 4-12, they went backwards, so 8am - 4pm week 1, then Mighnight to 8am week 2, and 4pm to Midnight week 3.
The limited research that has been done also suggests that rates of accidents don't increase if shift lengths increase from the traditional 8 to about 10, but do start to increase if extended to about 12 or more and this happens on a regular basis. However, difficult to find evidence that withstands statistical scrutiny in terms of the level of impact particularly in jobs that do not fall within types that have been designated "safety critical", often with targeted legislation to cover such jobs.
If all you want to do is count hours, then a decent timesheet system that is effectvely implemented (i.e. people are expected NOT to work hours but not record them!!) is probably most of what you need.
In contrast, if you want to manage "fatigue" then there are numerous variables to consider INCLUDING quite how bad the outcome could be if things go wrong.
....and by "quite how bad the outcome could be" I probably DON'T mean ONE fatality but rather multiple casualties and/or widespread environmental impact, and/or major impact on production and/or etc etc.
This is NOT to say that ONE fatality is OK, but all things are relative!! - for instance, very lttile tends to happen when there is a crash on the public highways with a single fatality in the UK - and even less if that were to happen in e.g. the US where the accident rates are MUCH higher. In contrast, all stops are rightly pulled out if a road bridge collapse and many people killed and injureed.