Ask a maintenance manager about the backlog and the answer is often a number of open orders: “about 400.” That says little. 400 orders of half an hour are a day’s work for a large team. 400 orders of two days each are a year.
A backlog is useful only when it is measured in hours and expressed in weeks.
The formula
Ready backlog in weeks = hours of all jobs with status “ready” / available hours per week.
In the example area, jobs worth 1,530 hours were planned and prepared. Eight technicians had 264 available hours per week. The ready backlog was 1,530 / 264 = 5.8 weeks.
Count only jobs that are ready. Jobs waiting for parts or approval belong to the total backlog, which is usually reported separately.
The healthy range
- Below two weeks: the team runs out of prepared work. Technicians start jobs that are not ready, or wait.
- Two to four weeks: healthy. The planner can build a full schedule, and jobs do not age too long.
- Above four weeks: jobs age. Small defects grow into breakdowns, priorities are overtaken by events, and the list loses credibility.
Why it was 5.8 weeks
The example area had three typical reasons for a high backlog:
- Preventive tasks for future weeks were entered as ready orders months ahead. These are planned, not overdue, and should be released only a few weeks before they are due.
- Old jobs that nobody wanted to delete: improvements without a sponsor, duplicates, jobs already done but never closed.
- Too few hours on planned work, because emergencies consumed most of the capacity.
The first cleanup removed duplicates and closed finished jobs. The rest came down as the planned share grew: twelve months later the plant-wide ready backlog stood at 3.1 weeks.
What to do when it is too high
- Clean the list: close what is done, merge duplicates, reject what nobody needs.
- Check the strategy: are preventive tasks generating work that finds nothing?
- Look at the reserve: if emergencies eat the schedule every week, attack the bad actors first.
- Add capacity only after the first three steps, temporarily through overtime or service providers.
What to do when it is too low
A low backlog sounds like good news, but it often means that notifications are not recorded or jobs are not prepared. Walk the area with a supervisor and look for defects that nobody reported. Check whether operators know how to raise a notification.
Watch the age
Next to the backlog in weeks, track the number of orders older than four weeks. A stable backlog with a growing number of old orders means that new work is always overtaking old work. Somebody needs to decide on the old jobs: do them, schedule them, or close them.
The next part turns to the parts that make jobs ready: spare parts for the critical assets.