4 ms·

Mars Helicopter Flight Delayed to No Earlier Than April 14

66 points by qwertox 6y ago

tectonic 6y ago
It took a little bit of work to get it there, we can wait a few more days.
boardwaalk 6y ago
The interesting bit: > During a high-speed spin test of the rotors on Friday, the command sequence controlling the test ended early due to a “watchdog” timer expiration. This occurred as it was trying to transition the flight computer from ‘Pre-Flight’ to ‘Flight’ mode. Though you can't tell too much from that. It could still be hardware or software and induced by the environment or not.
- sys_64738 6y ago
  systemd watchdog issue?
  - willyt 6y ago
    I’m not best qualified to answer, but some processors can have a watchdog timer that has to be reset by your software every so often otherwise within a set interval otherwise the chip resets. This works like a ‘deadmans handle’ that can be used to detect when your code has hung and will perform a hardware reset, so you’re system can return to a known good state. Might be something like that?
    - RogerL 6y ago
      It's dead common in embedded software, such as avionics. Every task needs a watchdog timer. Basically you 'ping' the timer every so often, which resets it to (say) 1 second (or whatever makes sense for your task) Then it counts down. If it reaches zero, meaning one second has passed since it was reset it will trigger an interrupt. You of course design your system so that your task trips it every 500 ms (say). You don't ever want that watchdog to expire. But, being human, we write infinite loops, or code that takes 'too long' to run. And so when the watchdog triggers the interrupt that is a signal that you have some kind of sw or possibly hardware issue. You shut down or restart that task, and/or do other remedial actions, like alert the pilot in human manned airframes. You of course have to be very thorough in your design here, as this is your last chance to react to these errors.
- 3v1n0 6y ago
  The text then mentions that there is no issue, so hardware seems safe. I guess there are self checks. Hopefully they have not forgot anything!
- canada_dry 6y ago
  Not clear which one though, as you can bet they'll have watchdogs for the watchdogs! One thing that's been made abundantly clear (once again) is just how cautious and methodical NASA is (not just when it comes to their $80M helicopter).
fudged71 6y ago
Aren’t they going to miss the 30 day window when the rover drives away? I remember them saying the rover is going to move on regardless of the progress/status of the copter.
- Denvercoder9 6y ago
  This is a 3-day delay, not too big of a deal. There's room in the schedule for unexpected problems.
jonahbenton 6y ago
Wonder if the flight control MCUs didn't respond in time. Will be fascinating to get the readout.
51Cards 6y ago
Still amazing to be living at a time when a non-fiction article starts with "Mars helicopter"
- simonblack 6y ago
  I found it amazing when I saw Sputnik fly over.
tails4e 6y ago
While I love the amazing engineering that goes into getting a helicopter onto Mars, the pace things are tried at is frustratingly slow. While I'm sure in part it's to ensure nothing goes wrong/not risk it, can someone speculate as to why it takes so long to get to a basic test flight? Surely all the SW and HW issues were well tested before launch, so why does it take weeks to be ready to fly after landing?
- aeternum 6y ago
  Agreed, I really hope this isn't the result of design by committee where each team adds a long list of tests because no one wants to be the one that didn't test something and lost the vehicle. In software teams, I've seen this issue of too many tests really cripple teams and result in frequent false-positives and products that never make it out the door. Hopefully that's not the case here.
- pengaru 6y ago
  Presumably simply letting time pass in every step's state along this process is gathering meaningful data. Jumping directly to "well we crashed the helicopter" would deprive you of answering questions like "how does the solar collector perform there?" "how are the heaters performing day after day?" "how does the helicopter fare resting on the surface throughout the varying conditions?" "Can we see a dust storm's effects on the airframe at rest and solar panels before flight maybe?" "how are the communications going? are they reliable in a dust storm?" That's just off the top of my head... Don't forget Ingenuity is primarily a tech demo and learning mission, they're just gathering data and informing future flight-centric missions. I think it's obvious that you'd leverage every pre-flight step as much as you possibly can, since flight has such a high risk of destroying the vehicle.
  - oalae5niMiel7qu 6y ago
    > "Can we see a dust storm's effects on the airframe at rest and solar panels before flight maybe?" This one in particular would have been kind of risky, since one thing that happens to airframes left to sit in windy conditions with no tie-down is that they get blown over.
simonblack 6y ago
My question about the helicopter is 'What happens if it falls from (say) six feet and lands on its side?' Does it have any way of standing on its feet again?
- londons_explore 6y ago
  If the rotor motors are strong enough, it ought to be able to turn them gently to get itself upright again. The motors might have insufficient torque for that though... Getting high torque on a direct drive motor means having lots of metal, which is obviously heavy...

66 points by qwertox 6y ago

It took a little bit of work to get it there, we can wait a few more days.

The interesting bit: > During a high-speed spin test of the rotors on Friday, the command sequence controlling the test ended early due to a “watchdog” timer expiration. This occurred as it was trying to transition the flight computer from ‘Pre-Flight’ to ‘Flight’ mode. Though you can't tell too much from that. It could still be hardware or software and induced by the environment or not.

sys_64738 6y ago
systemd watchdog issue?
- willyt 6y ago
  I’m not best qualified to answer, but some processors can have a watchdog timer that has to be reset by your software every so often otherwise within a set interval otherwise the chip resets. This works like a ‘deadmans handle’ that can be used to detect when your code has hung and will perform a hardware reset, so you’re system can return to a known good state. Might be something like that?
  - RogerL 6y ago
    It's dead common in embedded software, such as avionics. Every task needs a watchdog timer. Basically you 'ping' the timer every so often, which resets it to (say) 1 second (or whatever makes sense for your task) Then it counts down. If it reaches zero, meaning one second has passed since it was reset it will trigger an interrupt. You of course design your system so that your task trips it every 500 ms (say). You don't ever want that watchdog to expire. But, being human, we write infinite loops, or code that takes 'too long' to run. And so when the watchdog triggers the interrupt that is a signal that you have some kind of sw or possibly hardware issue. You shut down or restart that task, and/or do other remedial actions, like alert the pilot in human manned airframes. You of course have to be very thorough in your design here, as this is your last chance to react to these errors.
3v1n0 6y ago
The text then mentions that there is no issue, so hardware seems safe. I guess there are self checks. Hopefully they have not forgot anything!
canada_dry 6y ago
Not clear which one though, as you can bet they'll have watchdogs for the watchdogs! One thing that's been made abundantly clear (once again) is just how cautious and methodical NASA is (not just when it comes to their $80M helicopter).

fudged71 6y ago

Aren’t they going to miss the 30 day window when the rover drives away? I remember them saying the rover is going to move on regardless of the progress/status of the copter.

Denvercoder9 6y ago
This is a 3-day delay, not too big of a deal. There's room in the schedule for unexpected problems.

jonahbenton 6y ago

Wonder if the flight control MCUs didn't respond in time. Will be fascinating to get the readout.

51Cards 6y ago

Still amazing to be living at a time when a non-fiction article starts with "Mars helicopter"

simonblack 6y ago
I found it amazing when I saw Sputnik fly over.

tails4e 6y ago

While I love the amazing engineering that goes into getting a helicopter onto Mars, the pace things are tried at is frustratingly slow. While I'm sure in part it's to ensure nothing goes wrong/not risk it, can someone speculate as to why it takes so long to get to a basic test flight? Surely all the SW and HW issues were well tested before launch, so why does it take weeks to be ready to fly after landing?

aeternum 6y ago
Agreed, I really hope this isn't the result of design by committee where each team adds a long list of tests because no one wants to be the one that didn't test something and lost the vehicle. In software teams, I've seen this issue of too many tests really cripple teams and result in frequent false-positives and products that never make it out the door. Hopefully that's not the case here.
pengaru 6y ago
Presumably simply letting time pass in every step's state along this process is gathering meaningful data. Jumping directly to "well we crashed the helicopter" would deprive you of answering questions like "how does the solar collector perform there?" "how are the heaters performing day after day?" "how does the helicopter fare resting on the surface throughout the varying conditions?" "Can we see a dust storm's effects on the airframe at rest and solar panels before flight maybe?" "how are the communications going? are they reliable in a dust storm?" That's just off the top of my head... Don't forget Ingenuity is primarily a tech demo and learning mission, they're just gathering data and informing future flight-centric missions. I think it's obvious that you'd leverage every pre-flight step as much as you possibly can, since flight has such a high risk of destroying the vehicle.
- oalae5niMiel7qu 6y ago
  > "Can we see a dust storm's effects on the airframe at rest and solar panels before flight maybe?" This one in particular would have been kind of risky, since one thing that happens to airframes left to sit in windy conditions with no tie-down is that they get blown over.

simonblack 6y ago

My question about the helicopter is 'What happens if it falls from (say) six feet and lands on its side?' Does it have any way of standing on its feet again?

londons_explore 6y ago
If the rotor motors are strong enough, it ought to be able to turn them gently to get itself upright again. The motors might have insufficient torque for that though... Getting high torque on a direct drive motor means having lots of metal, which is obviously heavy...