THE GUIDANCE AM) NAVIGATION SYSTEM

The guidance and navigation system on board the Apollo command module was not only used for cislunar navigation exercises. It also formed an entire control system in itself that, to list just a few of its abilities, could be used to manoeuvre the spacecraft, control its attitude and make calculations relevant to many operations that might be carried out in orbit or during cislunar coast. It could fire the SPS engine, calculate the size and shape of orbits, aim cameras and other instruments at any target or maintain a desired attitude.

The design of the G&N was one of the first contracts awarded by NASA after President Kennedy set his lunar challenge. It was given to MIT, which had gained much experience in this field by designing inertial navigation systems for the US military for use in submarines, aircraft and the Polaris missile system. The Apollo design was based around three tightly integrated systems that worked together to provide a large range of functions.

How can you get to the Moon with just that?

At the core of the G&N system was the Apollo guidance computer. Now seen as primitive in comparison to its successors, it was nevertheless one of the items in the Apollo programme that helped to drive forward important technologies in electronics and computing. It demanded compactness and low power consumption, allied with high computational power – capabilities that could only be performed by a new device that was just coming out of American research laboratories: the integrated circuit or ‘chip’. When production of onboard computers for the Apollo programme was at its peak, it consumed fully half of the world’s output of integrated circuits yet only 75 units were constructed between 1963 and 1969. This was not because they were all used in the final machines, but because NASA bought vast numbers of the tiny devices from the manufacturers and hammered them with a barrage of tests to force ever higher quality control.

It is common for the Apollo guidance computer to be compared with modern domestic computers. More often, people display incredulity that a task perceived to be as complex as a mission to the Moon could be achieved with a machine whose computational power was, they believed, comparable to a digital watch, pocket calculator or other lowly item of electronic hardware. This is to misunderstand the nature of computers and how they work. Though limited in resources, the Apollo computer was carefully programmed at the machine code level. It did not require huge resources because its functions were very narrowly defined. The layers of abstraction that go into modern programming, where a high-level language has to be translated to a lower level of coding, were largely unnecessary; and computing power was not required to support sophisticated ancillary devices such as video displays. There was no word processor, spreadsheet, or even a simple decimal calculator.

The Apollo guidance computer (left) and a display and keyboard unit, the crew’s interface with the machine. (NASA)

Furthermore, its front end was not a QWERTY keyboard. Rather than make comparisons with modern stand-alone computers, the Apollo machine is better thought of as being like an embedded controller, tightly integrated into the spacecraft systems around it.

In hardware terms. Loo, it can be difficult to directly compare then and now. There was no one-chip processor at the heart of the machine. The processing unit was a card full of simple chips whose processing rate was 80,000 cycles per second, seemingly meagre in today’s terms. The data moved about the machine arranged as 15-bit words (plus a parity bit to detect errors), whereas computers from later generations settled on 8, 16, 32 or even 64 bits. Its sparse memory was very carefully and efficiently programmed with an extensive range of routines to assist the crews with the operation of their spacecraft. There were a total of 44 programs in the case of Colossus III – which was the name given to the software loaded into Apollo 15’s command module computer and this was packed into the equivalent of about 64 kilobytes of computer memory. This programming was stored on hand-verified, machine-wired core rope, an archaic memory technology that is no longer in use.

The crew ’spoke’ to the machine in a language of programs, verbs and nouns. Programs were numbered in groups according to the broad field of operation with which they were concerned. For example, programs used for the spacecraft’s descent to a planet’s surface were numbered in the range 61 to 67. Four programs for aligning the guidance system were numbered from 51 to 54. The selection of these programs and the functions they offered w’ere not arrived at easily. As Apollo went through its gestation, engineers, planners and crew’s wanted the computer to handle an ever-increasing range of tasks but, faced with its limited resources, they soon ran into difficulties. When programmers complained that the meagre memory available to the computer was filling up, their managers established an elaborate bureaucracy to carefully define w’hich tasks were essential and how’ best to achieve them, and left the rest off the machine. In truth, the computer was always running a number of programs simultaneously in order to carry out background tasks such as updating the state vector, but one program wns dominant at any one time, and w;as knowm as the major mode. The crew could call up a program as appropriate, or in some cases one program could call up another.

The crew’ gave the computer instructions using numerical codes called verbs. For example. Verb 49 was an instruction to automatically manoeuvre the spacecraft to a new attitude, and Verb 06 instructed the computer to display a set of three requested values in decimal form. Any value that the crew’ might wish to access was given a name, called a noun. Each noun was a numerical code that led to a value or a set of values stored in the computer. For example, during launch, the crew ran Program 11 in order to monitor their ascent. They punched in Verb 06, Noun 62 which asked the computer to display, in decimal form, three values that told them their speed, their height and how rapidly that height was changing. Internally, the computer handled the spacecraft’s guidance and navigation with metric units but because the crews were used to English units by virtue of their aviation background, it converted all relevant numbers to fect-pcr-second, nautical miles, pounds, etc.

A display and keyboard (DSKY) unit. This example was used after the Apollo programme to investigate computer fly-by-wire systems in aircraft. (NASA)

All interaction between the crew and the computer was by way of a dedicated display and keyboard, affectionately known as the DSKY and pronounced ‘diss-key’. This had ten numerical keys, a plus key, a minus key and seven other control keys that allowed the crew to engage in a dialogue with the computer. Above the keyboard were a cluster of lights to indicate the status of the machine and an arrangement of seven-segment displays, stacked vertically. Three of these displays, each with five digits, allowed the crew to see what data they were entering into the computer, or showed the result of the computer’s efforts. To keep the machine’s programming simple, there was no facility for the decimal point. Number entry and readout could be in octal (base-8) or decimal and was pre-scaled with the position of the decimal point assumed. It was left to smart astronauts to know where it should be.

Apollo crews came to respect the computer’s reliability and capability. David Scott said in 1982, "With its computational ability, [the computer] was a joy to operate – a tremendous machine. You could do a lot with it. It was so reliable, we never needed the backup systems. We never had a failure, and I think that is a remarkable achievement.”

This was not a stand-alone machine. It was tightly integrated into the spacecraft around it. It was linked to the optical systems and could both control and read the angles to which they were aimed; it could start and stop the engines; and it could

adjust the spacecraft’s attitude in relation to the reference that it gained from the gyroscopically stabilised guidance platform, i. e. its knowledge of which way was up.

The computer in the command module was called the CMC. for command. module computer. The lunar module had an essentially identical machine, the LGC or lunar – module guidance computer, which necessarily operated a different version of the software, named Luminary. Programming had to be specific to the tasks that were relevant to the spacecraft. For example, the LGC had to handle the lunar landing whereas the CMC needed routines for Earth re-entry. Also, the systems into which the computer was integrated were substantially different; for example, whereas the CMC only needed to start and stop the SPS engine, the LGC needed to control the throttle capability of the main engine for the descent to the Moon. They were not interchangeable.

The optics, described previously, formed the second part of the system in the command module. The sextant and telescope were not only useful for navigation, but being motorised they could also be commanded to sight on landmarks and track them to maintain the line of sight as the spacecraft passed overhead. The sextant’s optical power and tracking capability were such that a crewman in the command module could peer through its eyepiece and. if the coordinates were correct, see his colleagues’ landing craft on the lunar surface while passing more than 110 kilometres above at a speed of nearly 6.000 kilometres per hour. Moreover, he could Lake marks that allowed the computer to calculate the exact position of the lander.

The third major element of the guidance and navigation system defined which way w;as up.