# Could an Apollo LM land uncrewed?

I believe that Apollo mission control had the ability to upload data and commands to the LM's guidance computer as long as they had an operational communications link. I'm wondering if, from the point of separation from the CSM, there was any necessary procedure for landing which absolutely had to be done by an astronaut in the cockpit?

Some points of interest:

• After the braking phase of descent, the guidance computer would default into the P65 program, which was designed to do an automatic touchdown; all the Apollo mission commanders overrode P65 into P66, the semi-automatic mode, so they could choose the exact landing site. (Jim Lovell of Apollo 13 claimed he intended to use the automatic landing if it looked safe to do so, but never got the chance).

• The LM did the Descent Orbit Insertion (DOI) burn while out of contact with mission control; let's assume that either the flight plan changed to do two smaller burns just before and after going out of contact, or that the systems were modified to allow command of the DOI burn from the command module, or on a time-delayed command.

• The Powered Descent Initiation (PDI), and all subsequent landing maneuvers, occurred while the LM had line-of-sight to Earth.

I know that the lunar module pilot pressed a "Proceed" button to give the guidance computer the go-ahead to start burns of the descent engine; could Mission Control send the "Proceed" signal directly to the computer?

I know that the crew managed various mode switches on the way down; could the altitude radar mode be switched from Mission Control? Could commands be sent over the low-gain channel to adjust the pointing of the high-gain antenna?

I'm aware that P65 had no way of knowing if it was landing in a boulder field and there would be no very good reason to land an LM uncrewed, but I'm curious if it was "nearly possible", particularly in the context of a rescue landing.

• – a CVn May 24 '17 at 15:51
• Weren't many of the controls in the Apollo spacecraft (both the LM and the CM) physical rotary or toggle switches of various kinds? I strongly suspect those would require someone to physically manipulate them, but of course that does not preclude the possibility that they could be replaced with computer-controlled relays to let the computer more directly control the flight hardware. The computer would probably need to be augmented to do so, and of course the software would be different, but that's more in the realm of "nearly possible" than "nearly impossible"... – a CVn May 24 '17 at 16:17
• The core of my question is, would any of those purely physical switches absolutely need to be changed during the descent? Could a crew set the switches into a state that afforded all the landing-critical features, then retreat into the CM, prior to undocking? – Russell Borogove May 24 '17 at 16:27
• Apollo 12 demonstrated pretty good landing precision. According to Digital Apollo, P65 would have set it down about 300 meters from the Surveyor lander they were trying to reach -- the manual fly-down wound up landing about the same distance away in a different direction. Landing within 10km should be good enough for a rescue. – Russell Borogove May 24 '17 at 22:01
• @RussellBorogove: The automatic landing brought it to within a few feet of the ground just fine. They didn't know about the dust problem yet. I found the upward velocity documented long ago. The computer always tried to keep the engine bell pointed retrograde during descent; and was dumb enough to do so even when rising from the surface. – Joshua Sep 3 '17 at 4:16

I purchased "The Apollo Guidance Computer: Architecture and Operation" to answer this question (an excellent read). And the answer is yes it might be possible to land on the moon unmanned. However, a number of checklist items would need to be skipped, mostly numerous inertial system checkouts/alignments, most importantly for the gyros. They need to be within 2 deg to indicate a GO for landing (Page 262). The gyro alignment requires manual sighting of stars, something that cannot be done from the ground.

But ground control can make arbitrary DSKY (computer terminal) entries remotely$$^{[2]}$$ (page 340) allowing them to run the sequence of programs required for EDL (entry descent and landing). Assuming the gyros were calibrated before undocking, and had a drift rate similar to the one observed in the Apollo 8 Mission Report (Page 119) during the 48 minutes between undocking and touchdown there would have been $$(1\ mERU)^{[1]}(148\ min) = 0.04^\circ$$ of offset plenty good.

Here is roughly what the decent sequence would look like:

• Set AGS Mode switch to Auto
• Set PGS Mode switch to Auto
• Throttle Control switch to Auto
• (Set a couple more switches to their automatic positions, still cataloging the full set that needs to be enabled)
• Batteries 5 & 6 On
• Master Arm On
• Turn on the Helium tank heaters to bring the propellant tanks up to full pressure
• Final inertial system alignment update and crew exit
• Run TARGET DELTA V PROGRAM (P76) to perform orbital insertion
• Run Program 63 to start the landing phase (Titled: BURN, BABY, BURN -- MASTER IGNITION ROUTINE)
• Send remote KEY REL to indicate go for program execution, and send remote PRO, to start maneuvering in preparation for the descent
• Program 64 is automatically called to begin the Approach phase
• Program 66 is automatically called for automatic guidance and nulling horizontal velocity, slowly approaching the moon until touchdown
• It is possible that at this point Ground Control could poll the touchdown probes and issue an engine shutdown command when they see it, though I am unsure how fast they can poll, they could also just wait a set amount of time for the main engine to exhaust its fuel just kicking up dust.

Lunar landing procedure, from the Apollo 11 Flight Plan, sections boxed in red indicate IMU alignments or state vector updates:

[1]: mERU is a weird unit it is 1/1000th of the earth's rotation rate.

[2]: The bibliography contains more than a hundred references, and I am trying to find the reference that talks about this functionality.

• But a fully automated landing would require a flat plain with only very small boulders. – Uwe Mar 15 '18 at 9:04
• @Uwe that's true – Mark Omo Mar 15 '18 at 13:51
• Doesn't have to be a flat plain -- space.stackexchange.com/questions/17822/… -- and you can accept boulder distribution proportional to the degree of risk you want to take. – Russell Borogove Mar 15 '18 at 15:55
• Great answer! Some nitpicks and additional questions: Your footnote numbers appear to be reversed here. Is PRO (proceed) a DSKY function? I somehow had the impression it was a separate control, not part of the guidance computer, but I admit I have no idea why I thought that. Can you make the crew departure and LM separation explicit in your sequence of operations? At what point would the final pre-descent gyro alignment be done in the normal crewed sequence? – Russell Borogove Mar 15 '18 at 16:03
• @RussellBorogove I added an explicit crew egress, I could go into more detail as far as undocking, and switch placement (mostly all to auto, during normal operation they are not in auto to prevent glitches from wasting resources). I can highlight the times in the Flight plan where inertial checkouts and calibrations usually occur. – Mark Omo Mar 15 '18 at 16:51

Mark Omo's answer seems plausible. However, the official NASA answer was "no." A meeting at headquarters on February 11, 1969 about such a matter concluded that the spacecraft did not have that capability:

The possibility of an unmanned LM landing was discussed at NASA Hq. The consensus was that such a landing would be a risky venture. Proposals had been made which included an unmanned LM landing as a prerequisite to a manned landing on the moon. However, the capability to land the LM unmanned did not exist and development of the capability would seriously delay the program.

• I assume the major blocking issue was in guaranteeing a safe spot to land. Now I'm thinking about installing a camera at the commander's eye position and giving a ground controller control over the LPD redesignation in P65, which would at least give a fighting chance at it... – Russell Borogove Sep 22 '19 at 22:05