0
0

[–] flyingcuttlefish [S] ago 

These weapons entail a higher degree of risk than semi-autonomous (human in the loop) systems, such as C-RAM. The damage potential of the system depends on how quickly human operators can identify that the system is failing and take corrective action. In some situations, multiple unintended engagements could occur. Another crucially important factor in human-supervised control is the reliability of communications between the human controller and the autonomous system and what the system will do if it loses communications. If communications are lost with the human operator, will the system halt engagements

[from p.45]