Patriot and Aegis (the two major automated AD systems in the US that have seen combat) had three friendly/auto-related incidents, which were so different.
Patriot: Operators had little training and trusted the system too much
Aegis: Operators didn't trust the system at all
Patriot: Operators had little training and trusted the system too much
Aegis: Operators didn't trust the system at all
https://twitter.com/ThrustWR/status/1761936564874023012
The Aegis Combat System's doctrine stuff is like leagues ahead of what Patriot had in its autonomous mode. As the paper describes, it was "all-or-none." On the other hand, Aegis (due to distrust for the machine) had so many layers aka “decision leverage points” built in.
The system was extremely tunable from the start. Nonetheless, the human operators still made a grave mistake. In his book "Network Centric Warfare," Friedman states that the Navy later concluded that had the system been in auto, the Vincennes shootdown wouldn't have happened.
He says that is because for the system, Flight 655 wouldn't have met the internal threat/engagement criteria. But there's much to be said about these two distinct cultures that developed in the Navy and Army when it came to the machine.
The Navy's story starts with NTDS, it's worth reading about here: ethw.org/First-Hand:No_…
Anyways, if you look back on the history of combat systems, In attempts to overcome human reaction limitations, these systems were given autonomy, and in many cases, distrust in them by the operators has directly led to the deaths of many.
And in other cases, too much trust in these systems and a lack of “decision leverage points” has led to the deaths of some.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
