Mayday! Software Lessons from Aviation Disasters

Transcript

1. Mayday!
   Adele Carpenter
   Software lessons from aviation disasters
   @iam_carpenter
   

   View Slide

2. Photo by Graham Dinsdale via Jetphotos
   

   View Slide

3. Approximation of cabin
   Photo by Usien via wikipedia commons
   

   View Slide

4. Photo by Nick Bair via wikipedia commons
   

   View Slide

5. Chain of events
   Partnair 394
   

   View Slide

6. Photo by Chandler Cruttenden on Unsplash
   F-16 sonic boom?
   Lockerbie bombing Part II?
   Photo via wikimedia commons
   

   View Slide

7. Finn gathers the team...
   • How can we prove ourselves wrong?
   • What details might we be ignoring?
   • What else could cause a mid-air break up?
   • What are the precise levels? Other sources of residue?
   • No terrorist groups have claimed credit
   • Countless possibilities, need to narrow the scope
   

   View Slide

8. Photo by Chandler Cruttenden on Unsplash
   F-16 sonic boom?
   Lockerbie bombing Part II?
   Photo via wikimedia commons
   

   View Slide

9. Photo by Ana Municio on Unsplash
   

   View Slide

10. Question yourself
    • How can I prove myself wrong?
    • What details might I be ignoring because it doesn’t fit my
    theory or solution?
    • What else could cause this issue or situation?
    

    View Slide

11. Photo courtesy of BEA
    Photo courtesy of NTSB
    

    View Slide

12. View Slide

13. Photo by Brian Gordon via picryl
    

    View Slide

14. Photo by Nick Bair via wikipedia commons
    

    View Slide

15. View Slide

16. Photo by Brian Gordon via picryl
    

    View Slide

17. Chain of events
    Partnair 394
    

    View Slide

18. Counterfeit Code?
    

    View Slide

19. Improving software quality
    Finn’s top tips
    • Outcomes as a chain of events to learn from
    • Taking our time to investigate all options
    • Accepting responsibility our code - no “counterfeit code”
    • Following the evidence
    • Questioning ourselves rigorously
    

    View Slide

20. View Slide

21. Photo by Dan Parlante on Unsplash
    

    View Slide

22. Approximation of actual cockpit
    Photo by Keishi Nukina, KNaviation
    Captain McBroom
    First Officer
    Beebee Second Officer
    Frosty
    DC-8-61 Cockpit
    28 December 1978
    UA 173
    

    View Slide

23. Approximation of actual cabin
    Courtesy of Runway Girl Network
    

    View Slide

24. Approximation of actual cockpit
    Photo by Keishi Nukina, KNaviation
    Captain McBroom
    First Officer
    Beebee Second Officer
    Frosty
    DC-8-61 Cockpit
    28 December 1978
    UA 173
    

    View Slide

25. Photo by Randy Peters on Unsplash
    

    View Slide

26. Courtesy of the Oregonian
    

    View Slide

27. The cause was the Captain’s failure to properly monitor the
    aircraft’s fuel state and to respond to both the low fuel state
    and the crew-member’s advisories regarding the fuel state.
    The factor contributing to the accident was the failure of
    Beebee and Frosty to either to fully comprehend the criticality
    of the fuel state or to successfully communicate their concern
    to the captain.
    

    View Slide

28. Courtesy of The Flight Channel
    

    View Slide

29. Crew Resource Management
    Recommendations after UA 173
    • Participative management for captains: in other words
    providing a non-threatening environment that actively
    encourages collaboration
    • Assertiveness training for other cockpit crew members: how
    to get your point across effectively in a non-threatening
    manner
    

    View Slide

30. Courtesy of the Oregonian
    

    View Slide

31. Situational Awareness
    “Knowing what’s going on around you”
    Fixation -> loss of situational awareness
    

    View Slide

32. Situational Awareness
    Contributing Factors
    • Knowledge of the Automation and Abstractions present
    • Are you aware of the abstractions that you interact with, even
    if on a rudimentary level? Could you find the resources you
    need if the situation required it? Do you have a basic
    understanding of what happens when you run a Jenkins build?
    What’s it doing and what could go wrong?
    • Stress and Workload
    • How many tickets are assigned to you right now? How many
    meetings do you have planned? Which deadlines are real and
    which are arbitrary? Be honest with yourself and others on
    what you can accomplish
    

    View Slide

33. Situational Awareness
    Contributing Factors
    • Physiological factors
    • Did you sleep enough? Are you hydrated? Did you eat? Did you poop?
    • Abilities, Experience, Training
    • What are your limitations or weak spots with respect to your role? How
    can you patch them? Who can help you? What knowledge could you share
    with others?
    • Preconceptions and bias
    • Is your gut instinct actually just bias? Why does a certain course of
    action appeal to you? What evidence do you have to support your
    position? Evidence against?
    

    View Slide

34. Situational Awareness
    Is not a solo endeavour!
    • Contribute to a psychologically safe work environment
    • Share your knowledge, no matter your experience level
    • Speak up when things don’t add up
    

    View Slide

35. Thank you
    Adele Carpenter
    @iam_carpenter
    

    View Slide