When production's down, it's not the best time for doing anything not directly related to getting production running again (such as extensive testing or conducting repair history analysis). Yet when production is back up, there's usually another fire to put out so these things don't get done. Incomplete repairs, however justified at the moment, mean another failure too soon.
You can solve this problem administratively. First, ensure that fire-fighting mode is the job only of designated technicians. In a medium-sized plant, assign maybe three technicians to this role. Everyone else works uninterrupted on work orders (mostly repair and PM), so the work can be done efficiently and correctly. This seems infeasible when you're buried in emergencies, but not doing this is why you're buried in emergencies.
Second, ensure there's a work order for every repair that maintenance does. For emergencies, a supervisor can write the work order and bring it to the responding technician.
After the response tech turns in the work order, an engineer:
- Reviews the history on that equipment.
- Inspects the equipment and application.
- Makes a list of likely failure causes.
- Writes up a work order to investigate those causes, detailing which tests to perform.
Fight maintenance fires by removing their fuel.