By Alexander Fleiss:
The Boeing 737 max has been grounded since the second crash involving their faulty Maneuvering Control Augmentation System (MCAS). This system was designed to help the plane right itself during times of flight duress.
Both the Lion Air and the Ethiopian Airlines flights we’re doomed with faulty sensors that contributed to impossible situations for the MCAS system. The MCAS system was designed to work extremely specifically based on the data it was given by the computer, which is not dissimilar to any other algorithm.
Fact: A corruption of the data will lead to an immediate failure of the algorithm.
Whether the algorithm is for facial recognition economic predictions or as an auto pilot, this is an inescapable fact. Both the Ethiopian Airlines and the Lion Air have very poor track records for maintenance and upkeep of their airplanes. These are both relatively new airlines with very limited Government regulation and oversight as compared to what we are familiar with in the United States. Both of these airlines were expanding rapidly and were hiring very inexperienced maintenance crews. These maintenance crews did not fix the sensors for the two ill-fated 737s.
The flaw in Boeing’s design, because it’s the fault of both the airline and Boeing in the case of these crashes, was that the MCAS system wouldn’t ignore one of the sensors if the data came in very differently all of a sudden from the other sensor.
This is a common fail-safe switch which Boeing should have installed originally. Now this issue is being remedied with Boeing’s software engineers and has according to Boeing been fixed going forward.
So in the future, if the MCAS system gets two readings, one being incongruous with the other, the MCAS system will ignore the one that’s out of normal range for readings with what the MCAS would expect at that point in the flight.
With this change the Boeing 737 max should be airworthy. Yes, both crashes could’ve been avoided if the pilots had known to turn off the MCAS system, but at the same time if we are to push forward with automation in our society we must have safety features that allow for operators who might not have the skill level necessary. Computers and automation must have safety systems built in to overcome ignorant operators.
Alexander Fleiss is an experienced Chief Executive Officer with a demonstrated history of working in the financial services industry. He founded RebellionResearch.com, The Machine Learning Artificial Intelligence Robo Advisor, after 8 years working for mutual funds and hedge funds.