Fault injection testing is the practice of deliberately introducing failures into a robot in a controlled environment to verify that safety mechanisms work as intended. Instead of hoping failures never happen, teams actively test how their robot behaves when something goes wrong.
This is one of the most effective ways to prevent runaway incidents.
Real-world failures are rarely clean or predictable.
Common faults include:
If these scenarios are not tested intentionally, they often appear for the first time during a match—when the risk is highest.
Fault injection should cover both hardware and software failures:
Each test should confirm that the robot stops safely.
Fault injection should never put people at risk.
For every injected fault:
If the robot keeps moving, the safety system has failed.