AI & Technical question

Write an algorithm for a robot that has to get from point A to point B in a room.

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

What this question tests

Tests basic algorithmic and technical reasoning for a classic pathfinding problem, expected at a conceptual rather than full-code level for a PM.

Represent the room as a grid or graph, where each cell or node represents a traversable space and obstacles are marked as blocked.
Use a pathfinding algorithm like A-star or Dijkstra's, which finds the shortest path from point A to point B while avoiding blocked cells.
Explain A-star's key idea at a high level: it combines the actual distance traveled so far with an estimated distance remaining to prioritize which paths to explore first.
Address real-world complications: the robot needs live sensor input to detect unexpected obstacles not on the original map, and must recompute the path dynamically.
Mention practical constraints, like a minimum turning radius or the robot's physical size, which a simple grid model would need to account for to avoid unrealistic paths.

Names a specific, correct algorithm, A-star or Dijkstra's, rather than a vague find-the-shortest-path answer.
Explains the algorithm's core intuition simply and correctly, rather than describing it as a black box.
Addresses real-world dynamic obstacle handling, showing awareness this isn't just a static academic problem.

Describing the problem only in vague terms without naming any real pathfinding algorithm.
Ignoring dynamic, real-world obstacles and treating the room as a static, fully known map.

How would you handle a moving obstacle, like a person walking through the room?
How would you balance path optimality against computation speed on limited robot hardware?