Case context: A robotics team is training a warehouse camera system to decide whether a package is damaged and whether it should be routed for manual review. Instead of asking one model to solve everything at once, the team breaks the problem into stages: first detect the package outline, then find the shipping label, and then estimate whether the box has visible dents. They write these stages as explicit parts of the processing pipeline.

Question: Why does explicitly coding these subtasks help the learning system, according to the idea of task decomposition?

Sample answer: Splitting the problem into simpler stages and coding those stages into the system gives the learner useful prior knowledge. The pipeline no longer has to discover every intermediate step from scratch, because the structure already reflects how the task can be solved. As a result, the model can learn the overall inspection task more efficiently than if it were forced to learn everything end to end.

Key points:

• The complex task is divided into smaller subtasks.
• Explicit pipeline steps provide prior knowledge.
• Prior knowledge improves learning efficiency.

Rubric: The response must explain that 1. the task has been decomposed into simpler subtasks, 2. the explicit pipeline encodes prior knowledge, and 3. this prior knowledge makes learning more efficient.