HeapCrash / Design rounds / Task health monitoring and watchdog strategy

Task health monitoring and watchdog strategy

A FreeRTOS-based infusion-pump controller (non-safety-certified prototype) runs eight tasks: motor control at 1 kHz, sensor acquisition, alarm handling, UI, BLE, logging, a comms task to a host, and idle. Field units have frozen without resetting. Design how the system detects a hung or misbehaving task and recovers.

What the interviewer wants

Constraints

If you ask the interviewer

How it's graded

The rubric a senior interviewer would use. Each item is scored 0 to 3 and weighted.

Supervisor designweight 3
Liveness vs progressweight 3
Safe state and recoveryweight 3
Diagnosticsweight 2
Other hang sourcesweight 1
Validationweight 2

How to structure your answer

  1. Requirements and assumptions: the numbers that drive the design.
  2. Architecture: blocks, tasks, data flow, memory.
  3. Key decisions and trade-offs, with the alternatives you rejected.
  4. Failure modes and recovery.
  5. Testing and validation.

More design rounds

Write your design in five sections and get it graded against this rubric, with a follow-up question from the AI interviewer.

Answer this round