Case Study

Autonomous Delivery Robot Route Planning with Incomplete Inputs

Case context: You are building the navigation pipeline for an autonomous delivery robot. An upstream perception module provides nearby obstacle locations, but the route planner still produces unsafe and unstable paths. After checking the implementation, you confirm that the planning algorithm is working correctly for the inputs it receives.

Question: If the issue is caused by missing information, what is the most likely explanation, and what question should you ask to improve the system design?

Sample answer: The likely explanation is that the planner does not receive enough information to choose a safe route. Useful missing inputs might include sidewalk boundaries, curb cuts, or temporary construction barriers. To redesign the pipeline, ask: "What information, beyond the outputs of the earlier modules, would a careful human operator need in order to plan a safe route?"

Key points:

  • The available inputs are not sufficient for the task.
  • The planning method itself is not the source of the problem.
  • Ask what information a careful human operator would need.
  • Identify possible missing signals such as boundaries or barriers.

Rubric: The answer must diagnose the failure as missing input information and include the guiding question about what a careful human operator would need.

0

1

Updated 2026-08-12

Contributors are:

Who are from:

Tags

Machine Learning

Deep Learning

Supervised Learning

Dive into Deep Learning @ D2L

Data Science

Machine Learning Strategy

Machine Learning Yearning @ DeepLearning.AI