News Release - NASA Performs First Aircraft Accident Investigation on Another World
News Release - NASA Performs First Aircraft Accident Investigation on Another World
The main takeaway from the release to me is...
"Although Flight 72 permanently grounded Ingenuity, the helicopter still beams weather and avionics test data to the Perseverance rover about once a week. The weather information could benefit future explorers of the Red Planet. The avionics data is already proving useful to engineers working on future designs of aircraft and other vehicles for the Red Planet."
Paul, let me thank you again for all that you do to keep this community going. I know this question is kind of detailed, but I'd be happy if you could steer me even a little here.
I have some confused impressions about Ingenuity's last several flights. I seem to remember the flight team (or someone at JPL) mentioning that the drone's navigation software was having trouble orienting itself above mega-ripples and ripple fields, like the one occupying the Neretva channel, though Ginny had crossed plenty of ripple fields elsewhere in Jezero. If those fields were that disorienting, why was the team determined to fly the drone along that terrain, rather than directly across? (You can see that in flights 68-70, they didn't take the short way across the ripples, as they did in flights 36-40!) Some stretches of the Neretva channel do have steep sides, admittedly, especially as one moves in the upstream direction. I can imagine that they wanted Ginny to avoid that sloping terrain - well and good; why not follow the edge of the upper fan, then, alongside the channel, as Percy did? I do remember that the terrain was very blocky, and that it was slow going for the rover, but Ginny had navigated such terrain before itself, simply keeping up with the rover. They could have flown Ginny across the same relatively narrow and unrippled reach of Neretva that the rover took on the way to Bright Angel.
For me, the irony of Ingenuity's loss is that it did not occur in flying over all the variegated terrain of the crater floor - confusing even to geologists - or the steep cliffs of the delta front, or the weird surface of the upper fan... but in a ripple field. On a vertical hop, no less, when no lateral motion was planned. Geoscience me thinks, probably naïvely, that moderately-sized ripple fields like the one in Neretva are among the most organized and benign terrains the landscape offers in this part of Mars. They're not featureless like the smooth slopes of the crater rim - ripple crests are readily identifiable in Ginny NavCam images as sequential, distinct and curvilinear, forming high-contrast boundaries in most, if not all, cases. So you... land between them, where slopes are gentlest (and the drone didn't even seem to need the flattest slopes available!) I'm not a coder or engineer by any means, and I'm not trying to say that any of this is easy, but... if ripple fields are disorienting to the point where you must fly over them high and fast, shouldn't we have avoided them as much as possible?
I've had similar thoughts on this since the end of mission.
We knew early on in the mission that the springs on the legs had been reduced in weight and they were likely to be a failure point one day given enough flights. I assume that's why they seemed to favour soft landing on isolated ripples during the mission. But those were isolated ripples and caused no issues that we are aware of for the navigation algorithms. One point that was not discussed in the press release, but may come out it the end of mission report was the changes to the team size with team members being released onto other projects. I assume also some team downsizing was accelerated by budget issues caused by the success of the mission and the numbers of flights going way beyond those thought possible. Those team changes may have led to pressures to plan easy routes and accepting the risks by planning flights over those near featureless terrains. IIRC there were issues relating to localisation for the last couple of landings. In hindsight it's easy for us to say now, maybe after those localisation issues, then rather than plan a pop up flight to reset localisation, they should have flown up and away from the ripple field and pointed it in the general direction towards terrain with sufficient features for tracking and then permit it to decide where it was safe to land. I'm sure there will be more in the formal end-of-mission document. I hope it makes engineering and operational recommendations. I'll be watching for the release of the full report, and will share it here ASAP.
