The Real Story Behind the Apollo 11 Computer Error | WSJ


– [CAPCOMM] Houston, you’re
looking good for separation. You’re a go for
separation, Columbia, over. – [Narrator] On July 20,
1969, just moments after the Apollo 11 lunar module
began its descent to the moon a warning light flashed in the cockpit. – [Neil] 1202. – [Houston] 1202 alarm. – It’s a 1202.
– Standby. – [Narrator] The spacecraft’s
computer had overloaded and rebooted and no one knew why. As Houston scrambled to find an answer, an anxious Neil Armstrong
requested more information. – [Neil] Give us a reading
on the 1202 program alarm. – [Narrator] Back on Earth,
few were more nervous than a young computer programmer who had written the code
for the lunar landing. – We were landing on
the moon the first time. It’s not surprising there were problems. My name is Don Eyles, I wrote a good part of the computer code
for the onboard computer that was active during
the lunar landing phase of the Apollo mission. – [Narrator] Eyles career at
NASA began as a happy accident. – This was the summer of
1966, and I’d just turned 23. I was walking back home from a
rather dispiriting interview, I think at an insurance company. At that point I would have taken any job that anyone offered me. When I happened by the
MIT Instrumentation Lab and walked in cold and asked for a job. – [Narrator] He had never
written a line of code. Even so, Eyles was offered a position. That day he joined an army of over 400,000 scientists, engineers, and technicians working on the most
ambitious engineering project in human history. – No one knew how to land on the moon yet, just as no one knew how
to program the computer, and we would figured out both. – Hello, today we’re at the
MIT Instrumentation Laboratory, which has been given design responsibility for this guidance and navigation system, which will direct our Apollo spacecraft on the way to the moon and back. – [Narrator] In 1969, this
was the most sophisticated machine ever created. Before Apollo, computers had
mostly been vast behemoths, often taking up entire rooms
or floors of buildings. – When you say this computer is very much like land-based computers,
and yet I think of them as occupying whole bays of equipment. You’ve got all this
squeezed into a little box. How did you do that? – [Narrator] What was revolutionary about the Apollo computer was that it was the first use of integrated circuits, which allowed for a much
smaller and faster machine. – [Don] The computer was one cubic foot. It was roughly six inches
by a foot by two feet. And weighed I believe about 70 pounds. – [Narrator] Despite these advancements, the Apollo computer’s limitations presented formidable challenges. – [Don] We were dealing with a computer that was very limited both in terms of its memory capacity and its operation speed. What this book in front of me is is a listing of the flight code for the lunar module for
the Apollo 11 mission. This represents the contents of 36,000, 36k, words of memory. – [Narrator] To give you
an idea of just how small 36k of memory is, an average email message
today is about 75k. – Alternately for flight the
information in a book like this would be woven into a type of memory called core rope that was super reliable. The result of that was six
modules like the one in my hand. And these modules would be slid into slots in the back of the computer. And that would be the code. That was equivalent to plug in the CD-ROM into your early Mac. – [Narrator] Memory was so precious, the code that Eyles and
his colleagues wrote had to both do its job and also do so in as few characters as possible. – When you write a piece of code you’re writing something that needs to, in as few words as
possible, convey an idea. But at the same time it needs to fall trippingly off the tongue
of the central processor. You could call it a lapidary sort of art in the sense that you were
dealing with small things and trying to get them just right. (static hissing) – [Computer] 1202 alarm. – [Narrator] So what was happening during Apollo 11’s landing when
the computer was overloaded. – 1202.
– 1202 alarm. – [Houston] It’s a 1202, standby. – [Narrator] The computer’s
display was flashing error codes 1201 and 1202, but the astronauts didn’t know what those alarms meant. And for 50 years, neither
has much of the world. It all happened so fast. Not even the programmers who designed it were sure just what was happening. – [Reporter] There are many new things that are happening in this flight. There are big dangers involved, despite the best our technology can do and our technology does do very well. – Right here you see
the 1201 and 1202 codes. At the time, we were sort
of holding our breath. You know, what is going to happen next? Is the spacecraft gonna keep flying okay? Or is it gonna somehow go outta control? – [Announcer] They got a
momentary alarm on their system. – [Narrator] As the
spacecraft began it’s final descent to the moon, a terrified Eyles came to the conclusion that the mission was doomed. – There was a pit of the stomach feeling. If it had been up to me, I probably would have
recommended an abort. – [Narrator] But flight
controllers in Houston had a better perspective. Soon after the alarm started,
Mission Control realized that the computer was still running the critical guidance
and navigation systems. – [Houston] Go, same type, we’re go. Eagle, Houston, you are
go for landing, over. – [Narrator] Rather than
abort, they made the courageous decision for Apollo to proceed. Neil Armstrong took over
control of the craft and Apollo, of course,
landed safely on the moon. – [Neil] Houston, Tranquility Base here. The Eagle has landed. – There was no sense of blame. There was no one calling you and saying, “You fools, what did you do to us?” but there were questions being asked. It was up to us to figure
out what had happened. – [Narrator] The alarm issues were traced to an obscure condition in which a radar accidentally turned on,
flooded the computer with unnecessary data. – The alarms were saying
there’s no more storage space, we’re going to flush everything and sort of reconstruct it. Do what you would call a restart. – [Narrator] Eyles code
wasn’t bad, on the contrary, it had done exactly what
it was supposed to do. The issue was caused by
human error in the hardware. Someone had accidentally flipped a switch when it shouldn’t have been flipped. – It was determined that in fact the switches were set up in such a fashion that this weird condition could occur. – [Narrator] For his part,
Eyles stayed on at NASA and his code was deployed successfully in every Apollo mission. – What was developed was actually extremely advanced for the time and in some ways its more advanced that some of what’s being used
today in real-time systems. Because the greater
speed and greater memory of today’s computers don’t force you to be as compact as we had to.

100 thoughts on “The Real Story Behind the Apollo 11 Computer Error | WSJ

  1. I met and talked to this man at the maker faire 2019 in Rome and I didn't realize at all who he was …. I'm an ignorant!

  2. The alarms were caused because the computer was trying to process too much data because Aldrin decided to have the rendezvous radar on during the decent program. His thinking was to have it already running incase an abort was taken and they were to return to Columbia. It was poor a judgement call that caused it and not a malfunction. No other flights ever had this problem occur on their subsequent missions.

  3. This is not credible. A re-booting computer at the critical descent stage would kill the crew. A machine that under-powered (2K of RAM!) essentially leaves the astronauts piloting a high-speed missile headed to the Moon on fully-manual control, which is impossible. Landing a rocket vertically takes a massive amount of computational power, with adjustments being fed into the stabilisation systems in real-time. You cannot process this kind of raw data with a pocket calculator without crashing the system, which is what this pitiful machine was. Look at the travails SpaceX has had landing it's rockets on the ocean, blowing a dozen of them up in disastrously failed landing attempts, with computers with possibly a billion times the power of the Apollo machine? No, no, no.

  4. Let's not forget the brilliant people in charge of running the simulations for testing the mission control crew on this computer alarm in the last simulation before the real thing. If the simulation team hadn't tested it, mission control crew would have had no experience with it, and wouldn't have known how to respond to it. It was immense hard work of the whole Apollo project, turning over every rock, checking and rechecking everything, that made the project succeed.

  5. Eyles mis-programmed the flight computer and failed to pretest the installed landing sequence before launch. The matter was investigated and Eyles was suspended and sent back to MIT where he resumed his original job sweeping floors. She later litigated against NASA receive retirement pension. NASA won the decision. What happened after that? Who knows what these crazy people compromise for a pension.

  6. Have written assembler code with restrictions but nothing like their restrictions. The validation of the software must have been crazy extensive. The mind boggles how they handled change control.

  7. I read something written by one of the programmers that the multiple radar signals being processed would not have overloaded the computer if the multiplexer had sync-ed the signals properly. He laid the ultimate blame on the multiplexer for allowing unsynchronized signals to hit the IO processor. I believe the idea was that the IO processor was being interrupted twice as much as expected.

  8. The computer problem was even more serious on Apollo 14. They think ‘tin whiskers’ formed on a component causing a possible abort. I think they woke this guy up at MIT and he rewrote the code to prevent an abort, all while they were orbiting the Moon! The astronauts punched it into the DSKY and everything worked fine!!

  9. "The Real Story Behind the Apollo 11 Computer Error" To add drama to the fraud. What'd I win?!

    NASA. Apollo. Theranos.

    "Ignorance Is Strength." I'm weak.

  10. This video lost me in the first few seconds. The computer didn't "reboot". It went to the top of the priority stack once it had reached the maximum number of actions it could achieve in a cycle. The software was brilliantly written and saved the mission.

  11. Wrong it's because the radar could not lock onto the surface due to all the glass, the radar was having trouble finding the surface as the lander approached for lunar landing. NASA lies their a$$es off when it comes to the Moon & the things within it . Why do you think they referred to the Moon as the meatball? There's never been volcanoes or volcanic flows on the Moon hardly anyone realizes it's all glass & when we went to the Moon the astronauts walked on transparent ground. Down in the transparent ground of the moon are giant bodies massive ranging 500 miles in height, it's a ship encased in glass "alien glass" 3.4 billion years old that has collapsed permanently entombing the occupants!!! Using the correct lense filters 1 can see inside the Moon from Earth to see these giants that were driving the Moon. They've been preserved beyond belief it's the biggest story of everything. Take an image of the Moon rotate it upside down from the way it appears in the night sky and you'll see them. There are 7 visible that can be seen very well from Earth all inside the Moon buried under glass of amazing clarity. The apennine mountain range on the Moon is a giant wing of a white feathered giant, the feathers are miles long it's AMAZING!!!!!

  12. The programming for the landing sequence was written in error because the landing sequence was to be upgraded to the new language ,the programmer received the upgrade mod number but failed to comply with the reprogramming order however the mod revision number was changed indicating the change was made but no sign of a change actually was made. The programming codes were the same for a previous planned mission

  13. This was a bad, misinformed video. They have an actual expert on the issue behind those alarms and WSJ still managed to get it wrong.

  14. The rendezvous radar wasn't accidentally turned on; Neil and Buzz did it on purpose in case they had to abort. The radar took a little time to come online and they didn't want to risk the delay. The alarm when doing this showed up in the simulations with the Apollo 12 crew (prior to the 11 flight), it just didn't get shared. While I'm here, the "going long" was due to, again, Neil and Buzz deciding to keep the tunnel pressurized in case they need to scramble back into the CM. When they undocked the pressure release gave the LEM extra deltaV causing the mismatch in expected vs real.

  15. I don't know anything about computers, but That was Fascinating.. Now I understand my computer just a little bit better.

  16. To this day, I still have an old computer running DOS 6.22. I think that was the last version they put out. That computer has one, and only one task that it must do without delay or error. And it has worked flawlessly for many years. 16KB of RAM. 386 Processor. 1.4MB floppy disc drive. At work we used DOS for many years as well…..It didn't get confused trying to time share tasks. It stayed on point, on topic without fail. The Apollo computers are multi-tasking, but they were the same basic idea…solid as a rock and reliable.

  17. Why No humans can land on Moon
    Can some one answer this

    Every mission landing on moon failed with a Radio
    Are you serious ?
    Are we not alone in this universe

    Moon belongs to Indians .

    NASA Cheated

  18. Of course nasty would hire someone without experience, not that big of a threat inside a warehouse plenty of oxygen, air conditioning and film crew

  19. It's truly amazing how the stories keep rolling out. Suppressing the truth about the fake landings must be very exhausting for those involved. We were all taken for a trillion dollar ride, and they're not giving up!!!

  20. What I am missimg in this video is the fact, that Steve Bales and the backroom guys at MOCR have it figured out even before the mission started. In the simulations before the mission they were given the same type of alarm. When Bales called an abort, the simulation supervisors told him, that he shouldn't have done that. In reaction, Bales and his team, went over all the computer alarms and made a list of those, which called for an abort and those, which didn't. If you hear the audio from mission control during this alarm, You can hear Bales talking "Same type we've had" This refers to the alarms in the sim. So the guys in Mission control were pretty much prepared for this and did a perfect job in preparation and execution of the mission. "Tough and competent" as was the line from Gene Kranz after the Apollo 1 fire.

  21. It was Buzz Aldrins fault lol he had the “randevu” radar and “landing” radar on at the same time when procedure said one or the other…computer overload 😉

  22. State of art 36k memory on the Comp doing the most complex effort of that time! Thankx for the info. It's both humbling and imbues appreciation for all the hard work people did back than- laying the foundation stones, and then, the stepping stones for others to improve things later. Our lander just crashed on the moon 12 days ago with TB in memory and multifold of advances.

  23. LOL What nonsense! The problem was not real, it was a S I M U L A T I O N. Compare NASA's simulation videos with the alleged real videos and see if you can tell the difference.

  24. You guys have to be honored and respectful of these people that beyond any hope defeated the negative thinking of everyone else and decided to achieve the un achievable. That generation is almost gone , and we are getting stuck with the millennials that are happy with a job at McDonalds and a car .

  25. Pretty cool! Thanks for this post.  Do a post on what if Babbage had FLOWDAC tech, patent 3190554 , ( a digital computer that computes with air )  and had the pipe organ folks build his Analytic Engine for him.  Could be Lady Lovelace could have invent COBOL  , huh?   To Infinity and Beyond  !  Patents672256, 3013505 .

  26. In the mission report of Apollo 11, in the anomalies summary, read paragraph 16.2.5 (Computer Alarms During Descent).

  27. It’s like reading a rags to riches story like frank mccourt. He went through his famine to get to his grail. Oh I wish I was back in a time when elegant code was required. Now we have an army of copy pasters

  28. The coding Guy knew the operation was Doomed crashed, But NASA Astraunaut nothing to worry everything still working, offcourse everything works in a studio.

  29. Buzz also broke the switch for the engine to lauch the ascent back up to the command module. He jammed a felt tip pen into the switch to complete the circuit and lauch the engine.

  30. The funny thing is, in the thumbnail Mr Eyles does look quite a lot like the computer expert from The Office: https://www.youtube.com/watch?v=geZoES9KQ-Q "Wanna work for NASA? No thanks… I am making ** loads from Computers". BRAVO that man 😎🏆

  31. Lol, do some people still believe in the NASA fake moon landings? 🤣
    Science fiction brought to us by the forefathers of NASA,
    Wernher von Braun starting with his "Project Mars" , the bringing in Jack Parsons, Ron Hubbard, and Walt Disney to create the
    Science fiction fantasys 😆

  32. This doesn't tell the story at all! It wasn't human error (at least not in the LM, despite some of Buzz' claims). Murray & Cox explained that it was a failure of the change control process for the flight manual. This error meant the manual said to leave the rendezvous radar switched on when it should have been switched off.

  33. For those who are curious, there is a video of the Apollo 11's computer restoration on this Channel: https://www.youtube.com/watch?v=g0PIXvjTasI

  34. 6:10
    Incorrect , in fact , the errors were indeed caused by the randezvous radar being online , but it's Neil that let the breaker pushed in , so that in the case of a abort , the craft would imidiatly take the good trajectory. Engineers didn't think it would be activated , as in the checklist it was told to turn it off for descent.

  35. Even in the mid-90's, software engineering was heavily tied to the limit & cost of memory. I paid over $800 for 8 MB or ram and proudly bumped my 60 MHz Pentium up to 16.

  36. I wrote code for mainframe computers for 45+ years, and I'm pretty confident that not a single line of it is still running anywhere on earth.

  37. Oh… yeah… back in the day they (tech firms) hired people all the time that didn't know about computing… and if you had some experience it was an instant hire with better pay and more interestings assignments. If not, it would be just data entry more often than not.

    I used to (and still do) program embedded systems. Unknowingly that including early cruise missile terrain tracking back around 1979, although I was told it was for commerical aircraft applications. LOL! I had a brother in the Air Force doing maintenance of ICBMs, he had clearance, therefore I could not, so I was "on loan" to the military program from the commerical aircraft division. I knew nothing about terrain tracking. I did graphics and simulation, not telemetry.

    Anyway, everyone was hiring anyone they could that knew anything about computers and programming. Even Microsoft about that time was hiring right out of highschool.  

    I was (and some of my friends were also) hired away from college early by Boeing when they learned I (we) gain advanced programming skills on 32 bit IBM 360 TSO realtime system starting in 8th grade (age 13, 1972) as part of a Boy Scout's of America Explorer troop (yes Girls were there too) in Olympia WA, as part of a department of highways community and education outreach program started by a guy named Rich Glover (RIP).  

    Years later (around 1981), at Boeing, I leaned what the error codes meant on the Apollo 11 flight by reading the assembly code ourselves that someone managed to get a copy of. Out of memory error (buffer overflow) and the auto-recovery was to simply clear the data buffers and the realtime data feed refill it. No procressor reset. Just sound an alarm circuit and keep processing.

    We joked that the data toliet bowl got clogged and the program detected the overflowing data-turds and automatically flushed it. (which mean resetting the end of memory counter and index).

    Of course it would fill again and have to be flushed again after a while. We were pretty sure it meant that multiple data sources all dumping a stream of raw info into the data buffer from various input devices (like having more than one keyboard and typing on both and same with the mouse), thus maxing it out. But we had no idea what it was. The good part was that the data wasn't misleading. The guidance wasn't affected by the very different data values coming in, probably because it ended up in non-critical spots in the code. The other alarm was proably a calculation or data range error, which flushed the buffers too.

    This was standard indoctrination discussion at Boeing. I have no idea how Boeing got NASA code, but lot's of things got shared. My debugging tools showed up at NASA, DEC, HP, Rockwell, and other technical branches and services and I had no idea how. The middle managers across the industry often swapped companies and carring stuff with them. So I'd say it was pretty common code and other info got shared. There was some rudimentry internet (ARPNet) and email starting back then, and really getting going in late 70's.

    Fun stuff. I still do embedded systems, and trying to make things run fast and tight (low memory) is still an art.

  38. 4: 24 how can they have camera crews up there? Appolo has 2 capsule?, one capsule for the camera crew & other one for astronaut? who's the camera crew? can i be a camera crew? i hope can be "Stanley Kubrick"

  39. Flight trajectories calculated with 2k. Shuttle.commander complained they couldn't even do it for shuttle with desktops .
    2k computer . Zero radiation shielding, pass through the Van Allen's….lmfao
    Wake up Curverts !!!! Bwahahahahahahahahahahahahahahahahahahaaaaaaaaa .

  40. Do a google maps search for Bermuda Triangle . Then realise all NASA rockets trajectories head directly to it . Then realize the "myth" of the Bermuda triangle was created the exact same year as NASA. The perfect cover . Anyone research the Triangle "poof" lost in a vortex never to be seen again .
    Research it folks .
    Nasa Lies .

  41. " The Real Story " ? and you totally fail to mention Margaret Hamilton Nasa,s lead Apollo software designer who in fact designed and wrote in all those codes herself and whos genius & foresight made the landings at all possible, Don Eyes as a part of a team had input and a role sure but nowhere near as significant as hers. In 2016 Margaret Hamilton was awarded Americas highest civilian honor The Presidential medal of freedom from President Barack Obama for her pioneering software designs & successes on the Apollo project.

Leave a Reply

Your email address will not be published. Required fields are marked *