Corrective maintenance
Maintenance done after a fault is found, to restore an asset to working condition.
How it works
Something is wrong, and someone fixes it. The fault may be a breakdown that stops production, or a defect found during an inspection that has not caused a stop yet. Both are corrective work. The difference matters for planning: a breakdown is emergency work, while a defect found early can be scheduled like any other job.
Some teams use "reactive" or "run-to-failure" for the breakdown case and keep "corrective" for both.
Why it matters
Unplanned corrective work is the most expensive kind. Production stops, people drop what they are doing, parts are rushed in, and the repair is often done under pressure. A high share of emergency work is a common sign that preventive and condition-based maintenance are not catching problems early.
Corrective maintenance is not always wrong, though. For cheap, non-critical items with a spare on the shelf, running to failure can be the sensible choice.
In practice
A good corrective job ends with more than "fixed". The technician records the symptom, the cause they found, and the action they took. That record is what lets the next person fix the same fault faster, and what makes root cause analysis possible when the fault keeps coming back.
Time spent on corrective repairs is the basis for MTTR.