Loading page…
Challenge kit · 27 missions · 270 actions · 40 anomalies · 220 readings
Click any chart to see the records behind it.
Where SPOT's actions succeed, and the conditions that break them.
Below 25% battery only 33.3% of actions succeed (n=24), against 89.7% in the 61–80% band — the cliff the triage rules weight.
Indoor dusty waypoints are the weakest setting at 69.2% success across 13 actions.
V1 against V2, by action type and by the battery each mission spends.
Per-type success on each configuration. Small groups swing hard — hover a bar for its sample size.
A V2 mission spends 28 battery points on average, against 56 on V1 (-28 pts).
Every mission in time order, and what its reported status was worth.
Stacked actions per mission. A red cap on a mission that reported COMPLETED is a failure the status field hid.
21 of 26 missions that reported COMPLETED had at least one failed action inside.
Each bar spans start-of-mission battery down to where it finished. Bars that reach under 25% ended inside the cliff band.
What went wrong, how badly, and whether the measurement held up.
Reading unreliable is the leading cause, behind 14 failed actions.
Darker is more severe. High and critical findings are the ones the queue escalates first.
Every action as a point. Failed actions ran at 44% battery on average, against 58% for successful ones; duration barely separates them.
Share of each measurement the robot itself marked QUESTIONABLE — readings no one should quote as fact.
Failure rate at every stop on the route, split by the kind of work done there.
Turbine Hall South (high noise) fails 36.4% of its 22 actions — the worst stop on the route. Blank cells are pairings that never ran.