The space race of the 1950’s and 1960’s was one of the most pivotal events in the history of our planet. Mankind broke the bonds that held us on this world and set foot on another world. The best and the brightest of the United States and the Soviet Union competed to see who could control space and it could not have been done without the help of computers. As we get set for the anniversary of the Apollo 8 spaceflight, we examine one of the essential components that took the United States to the Moon and back.
It is a shock to many today that much of the calculations in the early days of space travel were done with a pencil and a slide rule. Computers at the time were nothing more than giant calculators and were incredibly expensive to build, operate and to maintain. The human brain was simply faster at the time and in many cases were women, most recently brought to life in the recent film Hidden Figures.
The Mercury Program took the first American into orbit in 1962 when John Glenn orbited the Earth months after Yuri Gagarin did. A primitive mainframe computer developed by MIT helped with the calculations that put Glenn and other Mercury astronauts into orbit. A computer plotted his trajectory in orbit but Glenn trusted Katherine Johnson’s calculations on the ground more than he did the machine. That would change.
The next prize in the race was to put a man on the Moon with the Gemini Program. This required complex calculations to be made and one wrong number could be fatal. Numbers had to be crunched quickly and correctly since an astronaut’s life could depend on it and a computer was the only thing that could do that. The problem was that they were too big. NASA gave IBM $26.6 million (about $216 million today) to build a computer with enough power to do what an astronaut needed but to also make it small enough to fit in the confined capsule that housed the crew. It had to be 19x15x13 and weigh no more than 60 pounds. This computer is roughly the size of modern day towers but would lack the multitude of engineers that were paid to keep the computers of the time running.
Docking with another craft was the first challenge to be tackled. The computer was needed to aid in this as it was a difficult and potentially deadly maneuver but one that had to be done if an American was to ever set foot on the Moon. The computer had to take into account the curvature of the orbit and the fuel consumption of the rockets to maximize the finite amount of fuel on board. Six different flight modes had to be programmed from pre-launch to reentry and that was just for a trip into orbit. It had to be sturdy enough to survive launch and the extreme G-forces that would be experienced going from a rest to nearly 17,500 mph and then survive the trip in the cold vacuum of space.
IBM was up to the challenge and devised a machine that understood only 16 different instructions but could execute 7,000 instructions per second. Magnetic tape stored information and there was a concern that errors from previous uses could creep into the system and be stored in the non-volatile memory. In space a flipped bit could be deadly. In those days it was acceptable to have an error every 100,000 bits but NASA demanded those specifications be an error every 1,000,000 bits. To ensure the correct program was loaded it would be loaded three times and would be checked for errors before being loaded into the computer. The Gemini Program’s guidance computer weighed about 59 pounds with core memory for 4,096 words and a 7.1 Khz clock. It could handle ascent, orbital flight, rendezvous and reentry and had a keyboard and display unit.
Once the feasibility of docking with a lunar lander in space was completed the next step was to go to the Moon itself. This was first done in 1968 when Apollo 8 circled the Moon. Apollo 10 followed in May paving the way a few months later on July 21, 1969 for Neil Armstrong and Buzz Aldrin set foot on the lunar surface with Apollo 11. They could not have gotten there without a computer.
NASA turned to MIT to design the guidance computer that would take the Apollo astronauts to the Moon. MIT had previously designed the guidance system for the Polaris Missiles so they were one of the few companies with the necessary experience. The system had to be programmed to account for the orbit of the Moon, landing and leaving the Moon to complete its calculations. An exact angle had to be achieved in order to enter the Moon’s atmosphere as well as the Earth’s. The astronaut’s lives were in the hands of a machine.
A risky new technology called the integrated circuit was used to reduce weight. Each chip had a circuit which only produced an output when it received a signal. Combined with others they formed a central processing unit (CPU). In the Apollo computer 5,600 integrated circuits formed the CPU. It ran at 1 Mhz, used 2.5 amps of power and had 36 KB of memory. It consumed 55 watts of power and weighed 70 pounds with a keyboard and a display unit as well. To help reduce power consumption a standby mode was programmed to use between 5 and 10 watts. The system was programmed using assembly code and the software was preloaded months in advance for early Apollo missions and could not be corrected in-flights if an error occurred. Both the command module and the lunar module had their own computers and the programming was so successful that it remained in service for use with the Space Shuttle, with several military aircraft and even the navy’s Deep Submergence Rescue Vehicle.
The programming itself was done by around 350 people with nearly 1400 person-years worth of effort. Two different sets of programming were used, one for the command module and one for the lunar module. The system watched for failures since a reboot was not possible in space but instead would reboot individual systems when they stopped responding. Perhaps the most famous instance of this was during Apollo 11’s descent when their radar overloaded the system that monitored it setting off the now-famous 1202 alarm at 6,000 feet above the lunar surface. The reset program kicked in and worked as it was designed to and a few moments later Armstrong and Aldrin were on the Moon. They stayed there for less than a day before returning to Michael Collins and the command module and returning safely to Earth on July 24 as national heroes.
In a letter afterwards Margaret Hamilton, the Director of Apollo Flight Computer Programming said “ If the computer hadn’t recognized this problem and taken recovery action, I doubt if Apollo 11 would have been the successful moon landing it was.” America had new heroes, Armstrong, Aldrin and Collins, flight director Gene Kranz and his crew as well. Shots of Mission Control replete with their computers and monitors were standard on every newscast around the globe. The forgotten heroes were the people who designed these computers, which could get a ship with three men to the Moon with no more processing power than a modern day calculator.