Francois Martin: One Small Test for Man, One Giant Leap for Software Quality
summary
NASA's 1970 report "What Made Apollo a Success" stresses testing as central: the 86‑page document has 271 occurrences of "test" versus 81 for development, and lists specific activities like 40 parachute drop tests. A table sums 4,104 hours = 2.33 person‑years of timed tests, teams correcting 65 failures per flight with fewer than two per vehicle from workmanship; better acceptance testing could have caught five of eight residual faults. Practices include early, environmental, redundant testing, heavy telemetry, vendor scrutiny (79 questions to subcontractors) and iterative mission development (Apollo 11 = 11th iteration).
AI-written summary · about
chapters (7)
- 0:00Apollo testing philosophy
- 3:38Testing for mission success
- 7:09Apollo acceptance testing
- 16:00Agile in Apollo process
- 20:40Apollo 1 fire lessons
- 27:03Apollo testing effectiveness
- 32:01Q&A
related talks
01h 13mRecipe for Testing Spacecraft Ground Software
44m30 years in testing – Battles won and battles lost
34mBalancing Early Testing with Real-World Reliability
44mMaking It Count: Quality is NOT an Option
21mMaking Software Testing Innovation Work
24mThe more we test, the worse?
56mThe End Game
26mThe Impact of Testing on a DevOps Pipeline