. . . Reaching for the Moon . . .
The LM Reliability Plan (page 315)

    The LM Request for Proposal (RFP) issued in August, 1962, emphasized that reliability would be a major consideration in source selection.

    Grumman management did not understand reliability technology thoroughly, like they did older engineering disciplines that they grew up with and in which they excelled.  They were extremely concerned, to say the least.  They were proceeding, to a great extent, based on faith in me. I was gratified, and felt a deep responsibility on my shoulders. With my unique training in reliability technology, I was the right person at the right place at the right time. This was a matter of pure luck.

    I felt strongly that a “reliability program” carried the first responsibility for assuring that the vehicle would perform satisfactory in its operating environment and that the manufacturing processes would be designed to assure compliance with design requirements. I had written and administered the reliability programs for the E-2C, F14, and OAO, and a number of other vehicles. From my activity on the Bureau of Naval Weapons-Industry Material Reliability Advisory Board (BIMRAB), I had observed what kinds of reliability programs the best companies in the US had, and I was convinced that my program was by far the best.

    When Grumman first received the RFQ, we noticed that no “mission analysis” had been written. Whatever information was available was scattered over many documents. So Grumman was awarded a contract to prepare a mission analysis for Apollo. My friend Tom was assigned the task and he asked me to provide him with the reliability requirement.

    In my group, we gave this matter much thought and came up with the theory that men will take risks, or can be asked to take risks, in proportion to the glory to be won by their achievement. To quantify the concept, I sent two men to the Indianapolis Race Track to determine the risk of death and injury. They came back with two numbers indicating the risks that men would take to voluntary win a dangerous international sports event.

    We thought that these numbers might be an American characteristic, and we wanted an international confirmation. I sent two men to Mexico City to determine the risks that Mexican Bullfighters take in the ring. They came back with numbers comparable to those that we got at the Indianapolis Race Track. We concluded that it is a universal human trait for men to take or accept certain limited risks in proportion to the glory that they can win.

    The glory earned by a man landing on the moon is greater than the glory that can be earned at the Indianapolis Race Track or the Mexico City Bull Fighting Ring. We can therefore ask such astronauts to take higher risks on a lunar excursion. Accordingly, I increased the risks measured at Indianapolis and Mexico City by an order of magnitude and established these numbers:

 Probability of Mission Success:        R=0.9840
 Probability of Crew Safety:       R=0.9995

    I wrote these numbers on a piece of paper and took them to Tom.  He just glanced at them, and thanked me kindly.  He inserted these numbers in his Mission Analysis, and they were accepted by NASA without further explanation.

The North American Reliability Program (Page 319

    After the submittal of the LM proposal, I was able to get copies of the North American Command and Service Modules reliability programs.  Each was a beautiful printed book, almost and inch thick, with an attractive light blue cover.  I studied them very carefully.

    I was appalled by what I read. The two documents were both the same.  They were written by the Advisory Group on the Reliability of Electrical Equipment and published by the Assistant Secretary of Defense. The North American documents were an accumulation of the highfalutin statistical theories that were being promoted by many pompous statistical reliability consultants that were running loose in those days.  Vehicle requirements were stated in Mean Time Between Failure terms that could not be demonstrated in a test and simply made no sense. There were no procedures to assure the reliability of the hardware. There was no demonstration program. 

    The review of these documents provided me with two tragic insights:

             1. North Americans did not understand what constituted an acceptable reliability engineering program.

             2. NASA did not have the capability to select a good reliability program.

    This was an untenable situation.  How was I going to sell my excellent reliability program to an agency that talked big about reliability, but had no idea how it was achieved? My last resort was to pray for a miracle.

Miracles Happen

    I gave a technical reliability paper at a meeting of the American Society for Quality Control in Las Vegas. One man in the audience kept asking a number of questions, and after the session, he came up front and asked me some more. He introduced himself as Jim Koppenhaven, Director of Reliability, NASA Headquarters, Washington D.C. I suggested that we go have dinner together.

    At dinner, Jim disclosed no statistical basis and appeared to be truly interested in what I had to say. Apparently, this was all new to him. After dinner, we went to a bar and talked until two am. I think I gave Jim a good introduction in reliability engineering and was relieved that I had had such a successful interview with NASA official. Miracles do happen