4 ms·

There is one thing a human can do that these new fancy drones can't yet. Look out the window and identify if a target is hostile or not. My understanding (I use

There is one thing a human can do that these new fancy drones can't yet. Look out the window and identify if a target is hostile or not. My understanding (I used to be in Naval flight training) is that this is still a huge part of the work done by pilots that these drones don't handle well.

Tloewald 13y ago
It previously struck me that an advantage of drones is that the pilots aren't under the time / personal danger pressure of needing to make snap decisions. In practice this turns out not to be the case, and -- as you say -- human pilots may make better judgment calls.
robryan 13y ago
If that is an issue surely the answer is more assisted drones. Better to have someone sitting in safety viewing a high def and possibly magnified camera output to identify things that someone taking a quick look in the cockpit while under fire.
- Jtsummers 13y ago
  Latency and mission. If you don't have line of sight with the drone, you've got multisecond latency on input, that's not suitable for fighter aircraft, though is not deal breaking for recon and air-to-ground (bombs or missiles). So fighter drones require the pilots to have line of sight to be effective in combat, which means you need control of ground or sea assets within range of the air combat theater. That's not a given in war. Though, at present, this is probably a given for the current activities of the US DOD, you can't base future planning on current conditions. That said, F-35 is a clusterfuck. Military contractors have had so little oversight, and are such terrible planners, (particularly of software) that most projects that depend on a significant software component are guaranteed to have cost and time overruns, with little repercussion except for the poor program manager that was thrown into the deep end and drowned.
nether 13y ago
NCTR (Non-cooperative target recognition) has advanced rapidly to the point that AESA radars can provide the geometry of a locked target and identify the type of aircraft BVR. There's also old-fashioned IFF.
rdl 13y ago
That seems like something which matters only in two situations: 1) Normal US style war, where we have complete domination of the skies and airwaves. No problem using high bandwidth comms back to base to get video in front of lawyers and flag officers to decide on the merits of a given target. False positives way worse than false negatives. (Also, for this kind of mission, even an F-18 is overkill; you could probably fly refitted F-4 or F-16 when the enemy is potentially a civilian airliner or whatever.) 2) "Real" war, where losing is a possibility (vs. just not winning). Comms may be disrupted. In that environment, the drone needs to be autonomous, but could make IFF decisions more loosely; within a bounding box, it's reasonable to assume any military aircraft which doesn't IFF as friend is in fact the enemy and can be shot down with no additional confirmation. Since the US prefers to engage BVR in that kind of situation, the drone has every bit as much data (from radar, emissions, etc.) as a manned pilot. False negatives and false positives are more comparable (maybe 10:1 vs. 10000:1 in severity).

Tloewald 13y ago

It previously struck me that an advantage of drones is that the pilots aren't under the time / personal danger pressure of needing to make snap decisions. In practice this turns out not to be the case, and -- as you say -- human pilots may make better judgment calls.

robryan 13y ago

If that is an issue surely the answer is more assisted drones. Better to have someone sitting in safety viewing a high def and possibly magnified camera output to identify things that someone taking a quick look in the cockpit while under fire.

Jtsummers 13y ago
Latency and mission. If you don't have line of sight with the drone, you've got multisecond latency on input, that's not suitable for fighter aircraft, though is not deal breaking for recon and air-to-ground (bombs or missiles). So fighter drones require the pilots to have line of sight to be effective in combat, which means you need control of ground or sea assets within range of the air combat theater. That's not a given in war. Though, at present, this is probably a given for the current activities of the US DOD, you can't base future planning on current conditions. That said, F-35 is a clusterfuck. Military contractors have had so little oversight, and are such terrible planners, (particularly of software) that most projects that depend on a significant software component are guaranteed to have cost and time overruns, with little repercussion except for the poor program manager that was thrown into the deep end and drowned.

nether 13y ago

NCTR (Non-cooperative target recognition) has advanced rapidly to the point that AESA radars can provide the geometry of a locked target and identify the type of aircraft BVR. There's also old-fashioned IFF.

rdl 13y ago

That seems like something which matters only in two situations: 1) Normal US style war, where we have complete domination of the skies and airwaves. No problem using high bandwidth comms back to base to get video in front of lawyers and flag officers to decide on the merits of a given target. False positives way worse than false negatives. (Also, for this kind of mission, even an F-18 is overkill; you could probably fly refitted F-4 or F-16 when the enemy is potentially a civilian airliner or whatever.) 2) "Real" war, where losing is a possibility (vs. just not winning). Comms may be disrupted. In that environment, the drone needs to be autonomous, but could make IFF decisions more loosely; within a bounding box, it's reasonable to assume any military aircraft which doesn't IFF as friend is in fact the enemy and can be shot down with no additional confirmation. Since the US prefers to engage BVR in that kind of situation, the drone has every bit as much data (from radar, emissions, etc.) as a manned pilot. False negatives and false positives are more comparable (maybe 10:1 vs. 10000:1 in severity).