"Approximately 54 seconds into the flight, a glitch occurred in the pipeline of images being delivered by the navigation camera. This glitch caused a single image to be lost, but more importantly, it resulted in all later navigation images being delivered with inaccurate timestamps."
These reports are well written and are a nice read, but more raw data from the flights and insights on how they do the analysis is what I long for. What happened in the image pipeline, how bad was the oscillations really. Being 5 meters off the landing is good for IMU flight, maybe...
This story has a very unusual level of meaningful content, presented with surprising clarity. The Ingenuity Project is an exemplar for how future space probe reportage should go.
I see two errors exposed. The first, attaching the wrong timestamps, looks like a simple coding error. The second, mixing navigational corrections into the closed-loop flight controls, is a much bigger booboo. It is to somebody's credit that the landing sequence, anyway, omits such input, although that might be just because they expected it to be going straight down. There is a right way to mix corrections into flight-control feedback, as an adjustment to calibration. But anyway navigational sampling also mixes in translation as a result of wind, which should not go into the control loop, but only into the goal seeking, at the next level up.
Somebody is learning a lot, in a short, stressful time, that could have been learned earlier from people who already knew it. Still, they probably won't forget it.
3 comments
[ 3.5 ms ] story [ 11.4 ms ] threadI see two errors exposed. The first, attaching the wrong timestamps, looks like a simple coding error. The second, mixing navigational corrections into the closed-loop flight controls, is a much bigger booboo. It is to somebody's credit that the landing sequence, anyway, omits such input, although that might be just because they expected it to be going straight down. There is a right way to mix corrections into flight-control feedback, as an adjustment to calibration. But anyway navigational sampling also mixes in translation as a result of wind, which should not go into the control loop, but only into the goal seeking, at the next level up.
Somebody is learning a lot, in a short, stressful time, that could have been learned earlier from people who already knew it. Still, they probably won't forget it.