top of page
Search
Writer's pictureAlex Kroll

Why we test



Boeing's recent Starliner failure was due to a software clock error which would have been caught in a fully integrated ground test. The test would have taken 25 hours, but the refly will cost $410 Million. See: https://www.space.com/boeing-defends-starliner-spacecraft-ground-testing.html.


The purpose of flight test is to reduce risk. We test to eliminate as many of the "unknown unknowns" as possible in a buildup approach to improve confidence in system maturity prior to advancing to the next level of test. This is a perfect real life example of the risk of not taking the appropriate steps to reduce the "unknown unknowns". Failing to eliminate risk leads to greater risks in the future.


While it's easy to assume that integration will be nominal after component level testing, the truth is that integration rarely goes as planned, and rushing can lead to expensive and risky mistakes. By failing to prove system maturity for launch, Boeing avoided a 25 hour ground test, but at the cost of a $410 million mission due to an overlooking coding error. Lessons learned in-flight are expensive. Far too many flight test programs have executed rigorous component level and analytical test only to fail once reaching the aircraft.

Don't get me wrong, these bench tests are incredibly valuable for determining component performance and building confidence and maturity; the truth is the more operationally representative environment you test in, the more you flesh out the "unknown unknowns". This includes the human in the loop.


Your program should execute as much test as you can afford, but bear in mind the technical, programmatic, and safety risk that comes with not building up to prove maturity at each step. Think small to big, and each level of test should be executed with the time and rigor to prove maturity for the next level. Be thorough, but efficient. Full integration testing is the last opportunity to ensure all components operate collaboratively prior to high risk flight test -- it is far cheaper to find problems at this stage than after leaving ground.


35 views0 comments

Recent Posts

See All

Comments


Post: Blog2_Post
bottom of page