The system keeps working correctly even when it receives invalid input or hits unexpected conditions.
Mental model: Like a car that shrugs off a pothole instead of losing a wheel.
Example: A CSV importer that logs and skips one malformed row instead of crashing the whole import.
In practice: Defensive programming — validating inputs, handling exceptions, and failing gracefully instead of crashing — is the day-to-day practice that builds robustness into code.