Please create an account to participate in the Slashdot moderation system

 


Forgot your password?
Close
typodupeerror
×
Moon

A Japanese-Made Moon Lander Crashed Because a Crater Confused Its Software (go.com) 37

Posted by EditorDavid from the space-mountains dept.
Last month Japanese startup ispace tried to become the first private company to land a spacecraft on the moon — but in the crucial final moments lost contact with its vehicle.

Now the Associated Press reports that company officials are revealing what happened: while trying to land, their vehicle went into free-fall. Company officials blame a software issue, plus a decision in December to change the touchdown location to a crater. The crater's steep sides apparently confused the onboard software, and the 7-foot (2-meter) spacecraft went into a free-fall from less than 3 miles (5 kilometers) up, slamming into the lunar surface. The estimated speed at impact was more than 300 feet (100 meters) per second, said the company's chief technology officer, Ryo Ujiie.

NASA's Lunar Reconnaissance Orbiter photographed the crash site the next day as it flew overhead, revealing a field of debris as well as lunar soil hurled aside by the impact. Computer simulations done in advance of the landing attempt did not incorporate the terrain of the new landing site, Ujiie said.

CEO and founder Takeshi Hakamada said the company is still on track to attempt another moon landing in 2024, and that all the lessons learned will be incorporated into the next try. A third landing attempt is planned for 2025.
This discussion has been archived. No new comments can be posted.

A Japanese-Made Moon Lander Crashed Because a Crater Confused Its Software More

A Japanese-Made Moon Lander Crashed Because a Crater Confused Its Software

Comments Filter:

  • It's unlike the Japanese to leave trash on the moon.

  • A crater? (Score:5, Funny)

    by nospam007 ( 722110 ) * on Sunday May 28, 2023 @04:20PM (#63557839)

    On the Moon?
    Nobody could have expected that, that would be like tigers in Africa.

  • We have all been there.. (Score:5, Funny)

    by zmollusc ( 763634 ) on Sunday May 28, 2023 @04:23PM (#63557849)

    You get to the critical last few kilometers and stupid fucking windows steals the focus to ask if you want sticky keys turned on and by the time you get control again, Jeb has died.

    • You get to the critical last few kilometers and stupid fucking windows steals the focus to ask if you want sticky keys turned on and by the time you get control again, Jeb has died.

      Or Windows 11 suddenly decides it is time to reboot to activate the most recent update that it quietly downloaded in the background.

      • Or Windows 11 suddenly decides it is time to reboot to activate the most recent update that it quietly downloaded in the background.

        ...which happens even when you bought the LTSB version of windows specifically so it wouldn't do that.

    • Not Windows' Fault (Score:5, Funny)

      by Deep Esophagus ( 686515 ) on Sunday May 28, 2023 @11:25PM (#63558393)
      They asked ChatGPT for the best landing procedure.

  • the 7-foot (2-meter) spacecraft went into a free-fall from less than 3 miles (5 kilometers) up

    Is it going to be okay, Uncle Larry?

  • 100% hindsight, I agree, but... (Score:5, Insightful)

    by greytree ( 7124971 ) on Sunday May 28, 2023 @05:02PM (#63557911)
    Surely someone in the loop would have said:

    "We ran the simulation for the old landing site. We just changed the landing site, so ... shouldn't we run the simulation for that?"

    Surely?

    • Somebody should have said that, sure. Seems like one of the following happened:

      • Nobody did, or they were afraid to speak up to 'authority'? (seems to be a thing in Japanese culture).
      • Somebody did, but it wasn't picked up by others.
      • It was proposed, but management brushed it off.

      In other words: might be more of a people / management / communication problem, rather than an engineering issue?

      • "The software was provided by US space software developer Draper", according to the report. Sounds like a last minute change didn't get communicated widely enough.

      • ...they were afraid to speak up to 'authority'?

        In 1893, an entirely predictable collision between two Royal Navy ships happened due to Vice Admiral Sir George Tyron's order to perform an unconventional maneuver. Bridge officers on the Admirals ship realized Tyron's error immediately but remained silent. The commander of the ship which would ultimately ram and sink the admirals flag ship (with 350+ lives lost) also recognized the inevitable result of this order. He delayed his turn but was was chastised by the Admiral, and so, followed his orders

    • Re: (Score:1)

      by Anonymous Coward
      Or most likely, they did run many simulations (remember, they had partnered with Draper, who are not exactly new to the space industry) but the simulations did not model accurately enough that specific part. Creating and running accurate space environment simulators takes a *lot* of resources, human and computers.

    • Re:100% hindsight, I agree, but... (Score:4, Informative)

      by khchung ( 462899 ) on Sunday May 28, 2023 @08:33PM (#63558169) Journal

      Surely someone in the loop would have said:

      You forgot this was made in Japan, Japanese in the loop would never rock the boat. It is much easier to go with the flow first, and then to apologize and bow afterwards. If one bow is not enough, then bow again.

    • Why does it even need a simulation? Radar altimetera have been a thing since WW2. Sounds like they decided to get clever with the landing system and it bit them. KISS is something many engineers forget.

      • It is hard to work out what will break in a system that you have optimised, because optimising usually make the system more complex.

        It is easy to be sure that when you change something, you retest it.

        Doing the latter would likely have avoided the former dooming this mission.

    • In Japan you do not talk back to your boss.

    • So basically, an abrupt change in measured altitude (due to an actual cliff) caused the software to stop trusting its altimeter, and it did its best to measure altitude by dead reckoning instead, which of course didn't work. This is why redundancy is important.

  • The estimated speed at impact was more than 300 feet (100 meters) per second, said the company's chief technology officer, Ryo Ujiie.

    Aha. There's your problem, right there.

  • "Computer simulations done in advance of the landing attempt did not incorporate the terrain of the new landing site"

    Cost cutting, I guess. Simply amazing.

    • Re: (Score:2)

      by cmarkn ( 31706 )
      I think it's more likely a scheduling issue. Management was stuck on a delivery date that didn't allow for retesting with the new landing site, only for the changes in the software.
      Either way, it's a management issue, not an engineering issue.
    • Left hand doesn't know what the right hand is doing. Film at 11.

  • The crater didn't confuse its software.

    The altitude measured by the instruments didn't match the altitude used by the lander's onboard maps. It then determined that the instruments were faulty, and used the onboard map data.

    Unfortunately, the onboard map was for the previous landing site.

    There was no confusion, just bad data due to a poor process.

    Who comes up with these shit summaries?

    • Reminds me of the worst Antarctic disaster [wikipedia.org] where a tour plane with 237 people onboard had its programmed flight plan changed without notifying the flight crew. When a "flat white" condition occurred (not rare in the arctic) and the crew could no longer see what was in front of them, they did not know that the flight computer was flying them straight into Mount Erebus.

  • It didn't crash, it proved its value as a kinetic weapon for lunar warfare ;-)

Slashdot Top Deals

Real Programmers think better when playing Adventure or Rogue.

Close