A SMALL STEP FOR COMPUTERS, BUT A GIANT STEP FOR MANKIND
Are we ready to trust computers? Neil Armstrong wasn’t and it saved the Apollo 11 team, Robert Matthews writes
From crash-landing a hypersonic aircraft to bringing a wildly spinning spacecraft back under control, there’s plenty of action in First Man, the blockbuster biopic of astronaut Neil Armstrong.
At one point Armstrong, played by Ryan Gosling, faces so many challenges during his historic landing on the Moon that it all seems a bit far-fetched.
Computer alarms are going off, the radar keeps bugging out, there is virtually no fuel left and Armstrong still cannot find anywhere to touch down safely. If it is true that the movie bends the facts, the reality was much worse.
Mission Control kept losing contact with the lunar module, which was so distant that every communication took 2.5 seconds to make the round trip. The movie brilliantly conveys the audacity of Armstrong’s achievements, all made using technology that today is more than half a century old.
But the film also holds lessons for those who think 21st century technology is smart enough to trust. By today’s standards, the computing power used to make Armstrong and Buzz Aldrin the first men on the Moon is risible.
A cheap pocket calculator packs a bigger punch. Certainly it could have spared the two astronauts the alarms that kept flashing during the descent, which were the result of the primitive on-board computer being overloaded.
But even today’s computer technology can be undone by a problem that nearly killed Armstrong and Aldrin – Gigo, or garbage in, garbage out.
It’s at the heart of a mystery that the movie dodges. How, despite all the planning, practice and dry runs, did Armstrong and Aldrin still end up staring disaster in the face?
The Apollo programme cost more than $200 billion (Dh734.64bn) in today’s money and was arguably the most complex engineering project undertaken. It was also unprecedented in the determination of those involved to think of everything.
From using fuels that did not need ignition systems to predicting deadly radiation storms from the Sun, America’s brightest minds dealt with every conceivable threat … almost.
As the module came into land, Armstrong saw the landing-site was a huge crater with car-sized boulders around it. This was not in the plan.
Mission planners had spent years pondering the safest place to make the first manned landing. They scoured the best Earth-bound telescopic images. Then they sent low-flying craft over the Moon, which were able to take images of objects as small as a metre.
After one last check using images from Apollo 10 weeks before the historic mission, the perfect spot was found on the Sea of Tranquillity. It was well lit, dead flat and clear of obstacles.
So how did everyone miss the huge boulders and crater the size of a football field?
They did not. Even before the final approach, Armstrong had noticed they were shooting past landmarks earlier than expected.
For some reason, the module was travelling faster than anticipated, and had overshot the planned landing site. But the on-board computer was doggedly sticking to the original flight-path.
Realising that disaster loomed, Armstrong took over the controls and began a desperate search for a clear site.
After several anxious minutes, with computer alarms flashing and barely any fuel left, Armstrong found a spot and pulled off a perfect landing. The rest is history.
But Nasa’s mission controllers knew tragedy had only narrowly been averted and set up an urgent inquiry. With
Apollo 12 scheduled for launch just a few months later, they had to discover what they had missed.
The culprits were two subtle but potentially deadly effects that no one had considered.
The first was the result of the module detaching from the mother ship to make its descent. As they separated, air seeped out from the tunnel that had connected them. This gave the module a light push and nudged it off the expected trajectory.
The second culprit was hidden beneath the Moon’s surface.
The year before the Apollo 11 mission, astronomers found evidence of colossal chunks of rock buried deep in the Moon. Their origin was unknown but their huge mass distorted the gravitational field of the Moon and pulled the module dangerously off track.
By the time Apollo 12 went to the Moon, Nasa’s engineers had found solutions so effective that the module landed just metres from its target.
But the fact remained that they had tried to think of everything that could possibly go wrong and had failed. The reason Apollo 11 succeeded was because Armstrong realised something was wrong and seized control from the computer.
That was only possible because the engineers had a backstop for their lack of omniscience. They did not feel comfortable giving computers total control of the descent, so they gave the astronauts the option to take over.
Armstrong proved the wisdom of that decision, and later missions made it even easier for the astronauts to seize control.
But half a century later, engineers seem to have again fallen under the spell of computer power.
Just last week, car makers General Motors and Honda announced plans to work together on a vehicle with “Level 5 autonomy”. That is, humans do nothing apart from get in and give the destination. There is no steering wheel, pedals or gear stick.
We can only speculate what Armstrong, an engineer and test-pilot turned astronaut, would have made of such confidence in computers.
Perhaps the makers of such cars truly believe they can do better than America’s finest did on Apollo, and really can think of everything.
Half a century later, engineers seem to have again fallen under the spell of computer power