3 ms·
Can someone chime in on the following-- One possible lead in cause has to do with either software or sensors related to Angle of Attack (AOT). For such an inte
by ineedasername 8y ago
Can someone chime in on the following-- One possible lead in cause has to do with either software or sensors related to Angle of Attack (AOT).
For such an intermittent bug in a safety-critical application, how do you even begin fixing it when reproducibility might be a 1 in 500,000 likelihood? (Obviously I'm not in software testing, but I'm curious how one approaches this problem.)
- peteradio 8y agoGet somebody who isn't a shit head to review the design. I bet it sucks the big one in ways that would be apparent to a real professional.
- rfugger 8y agoUsing flight data from two crashes?
- djsumdog 8y agoThere's another comment on how Southwest ordered retrofits of the AOA user interface and training for all pilots following the 2018 crash. One solution is to make that change global in a mass recall/retrofit, along with additional training; possibly even adding it as a situation in 737 simulators to judge how quickly pilots will react to it.
- Abott 8y ago"Metamorphic testing of driverless cars" is a very related article involving software testing and sensor problems for safety-critical autonomous vehicles. The article is available here: http://dx.doi.org/10.1145/3241979 http://dx.doi.org/10.1145/3241979 and a free copy is here: https://www.researchgate.net/publication/331289445_Metamorphic_testing_of_driverless_cars https://www.researchgate.net/publication/331289445_Metamorph...