Your question is Design an Edge Case Mining Pipeline. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You are building a pipeline that scans autonomous driving fleet data to find rare edge cases, such as construction zones or unusual emergency vehicles, and route the best examples for human labeling. The goal is to improve downstream perception and prediction models without overwhelming labeling operations.
How would you design a data ingestion and active learning pipeline to identify and label rare edge cases such as construction zones or unusual emergency vehicles from fleet data?