Drone Pilot Shortage And The Path To Machine Autonomy

A recent story in The Daily Beast reports on an Air Force memo that warns the US drone fleet is at "the breaking point." The Air Combat Command, which handles training of Drone operators, is failing to provide an adequate number of pilots for the increasing number of requests by the combat commanders for drone support.

Here are a excerpts from the memo:

“ACC believes we are about to see a perfect storm of increased COCOM [Combatant Commander] demand, accession reductions, and outflow increases that will damage the readiness and combat capability of the MQ-1/9 enterprise for years to come,” reads an internal Air Force memo from ACC commander Gen. Herbert “Hawk” Carlisle, addressed to Air Force Chief of Staff Gen. Mark Welsh. “I am extremely concerned.”

“ACC will continue to non-concur to increased tasking beyond our FY15 [fiscal year 2015] force offering and respectfully requests your support in ensuring the combat viability of the MQ-1/9 platform,” Carlisle added.

[...]

“The reduced offering of 62 CAPs (plus a 60-day Global Response Force) has been submitted to the Joint Staff; however, the Joint Staff has indicated their desire to circumvent normal processes while proposing their own offering of 65 MQ-1/9 CAPs,” Carlisle wrote. “This simply is not an option for ACC to source indeterminately.”

[...] >p? “ACC squadrons are currently executing steady-state, day-to-day operations (65 CAPs) at less than an 8:1 crew-to-CAP ratio. This directly violates our red line for RPA [remotely pilot aircraft] manning and combat operations,” Carlisle wrote. “The ever-present demand has resulted in increased launch and recovery taskings and increased overhead for LNO [liaison officer] support.”

It gets worse. The Air Force has been cutting short training of new operators to keep drones in the air and has been doing such career and morale damaging things as canceling training from existing operators and cancelling leave. This is leading to operators leaving the service.

It is not clear how much this problem at the front-end (among the pilots) is reflected in back-end operations (e.g., support and maintenance personnel, intelligence officers to analyze the information the drones are producing). However, it is doubtful that his high level of operations tempo is not negatively affecting all aspects of the US drone fleet.

Historically, it is not a good sign when a military is cutting training or sending trainers into operations. During World War Two, for example, the Luftwaffe's training program completely collapsed as the war progressed, with training time cut and senior pilots sent into combat, rather than passing on their skills and experience to the next generation of pilots.

If we cannot correct this problem by increasing the number of operators being trained and making certain that each operator is fully qualified, then that will be even more of an incentive to move towards autonomous systems. The problem we are facing here is not about platforms; if each of these could fly without any human intervention, beyond giving it an initial mission, then manpower could be focused more on support and battle management.

We are already moving in this direction. Any manpower crisis that impacts the ability to conduct operations in the field will only accelerate the process. And it should. Autonomous systems present new challenges and vulnerabilities; but the benefits outweigh these costs. The biggest benefits are the removal of Americans from harms way and the reduction in the number of people needed to maintain a robust defense. Given that manpower costs are one of the biggest items in the defense budget (and entail lifetime spending on pensions and medical care) reducing this is critical to developing new technology and procuring and supporting weapon systems.

There is an odd bias against autonomous combat systems. I understand why some military personnel are opposed. If I pilot an F-22, I don't want to be told that the next generation of fighters won't need me. But, I think most of the opposition comes from people who have visions of Terminator in their head. Or, they think that using machines will make it easier to use force. The former fear is unlikely. An autonomous bomber, for example, does not need to be intelligent; it just needs to be able to conduct it's mission. There is no way for a system like this to suddenly decide it hates humans or wants to do some other off mission action. As for the latter, this may be true. However, that means people have to be willing to hold their leaders accountable, not that we reject a promising technology.