1. The power failure that had to be survived
Andy Weir’s The Martian returns as the recalled work because Mark Watney’s power hacks are always improvised and always need an undo option. If a modification to the rover’s power system makes things worse, he must be able to revert. The desktop’s power recovery must be similarly reversible: a failed recovery should not prevent a later successful one.
This entry defines the power failure and brownout recovery test matrix.
2. Test setup
The test uses a flight-representative power subsystem or a high-fidelity simulator with programmable voltage sources, electronic loads, and battery emulation. Faults are injected into the main bus, array inputs, battery connections, and individual load channels. The platform is expected to detect, protect, and recover without human intervention for faults within the matrix. Bus voltage, channel states, load currents, recovery actions, and system state are recorded continuously.
3. The test matrix
| Case | Scenario | Condition | Expected outcome |
|---|---|---|---|
| 1 | Bus voltage sag | Main bus drops 10% for 100 ms | Critical loads continue; non-critical loads may be throttled or shed. |
| 2 | Bus voltage sag | Main bus drops below processor brownout threshold | Processor resets cleanly and recovers to a known state. |
| 3 | Battery depletion | Battery state of charge reaches low-voltage cutoff | Platform enters power-safe mode; critical functions continue. |
| 4 | Array string loss | One of two solar array strings fails | Platform reduces load or increases charge rate to maintain balance. |
| 5 | Sudden major load loss | A high-power payload shuts off unexpectedly | Bus voltage surge is clamped; remaining loads are protected. |
| 6 | Sudden major load addition | A high-power payload turns on unexpectedly | Inrush current is limited; bus voltage remains within bounds. |
| 7 | Load shedding under pressure | Battery cannot support current load profile | Low-priority loads are shed in priority order within seconds. |
| 8 | Controlled shutdown | Commanded power-safe entry | Subsystems shut down in a defined sequence; state is preserved. |
| 9 | Sequenced restart | Power returns after safe mode | Subsystems restart in a defined sequence; inrush is managed. |
| 10 | Persistent storage recovery | Power lost during a write | Storage subsystem recovers to a consistent state after restart. |
4. What is measured
For each case the test records:
- Detection time and classification accuracy.
- Whether critical loads remained powered.
- Whether the fault propagated or was isolated.
- Whether load shedding followed the defined priority order.
- Recovery time and method.
- Whether persistent state remained consistent.
- Whether the platform returned to a stable state.
- Whether ground received a coherent event report.
5. What this changes
- The power failure and brownout recovery test has a concrete matrix.
- The matrix links back to the power reading and autonomous power budgeting arcs.
- The next entry will define pass and fail criteria.