Reading between the lines

Air-to-air UCAVs unlikely in the near future

mq1_baladMQ-1 Predator from 332d Expeditionary Air Wing takes to the air over Joint Base Balad, Iraq.  (U.S. Air Force photo/Senior Airman Tiffany Trojca)

Outgoing Pentagon acquisition czar John Young had a parting shot for the Air Force, criticising the service for not moving quickly to implement an auto-land feature which might help reduce the number of UCAV losses due to signal delay or human error.  I think AFA’s account captures it best, because it also includes interesting anecdotal information:

Cherrypicking Data: On Monday, John Young, in his final moments as the Pentagon’s top weapons buyer, levied criticism against the Air Force for not doing more, as the Army has done, to incorporate an auto-land capability with the MQ-1 Predator unmanned aerial vehicle. He cited this as a leading cause of Predator attrition. Here’s what Mr. Young didn’t say: The Air Force has been focused, per Office of the Secretary of Defense guidance, on getting as much overhead intelligence-surveillance-reconnaissance capability to the war theater as possible. Today, USAF provides 35 simultaneous combat air patrols of Predators and MQ-9 Reapers to support the fight in Iraq and Afghanistan. It is on track to supply 50 CAPs by 2011, two years before the Army’s own version of the Predator, the MQ-1C Sky Warrior, is ready for combat. Yes, Sky Warrior will have an auto-land capability and the Air Force may very well adopt it when it is mature. But, for now, there is no auto-land system available for Predator. (The auto-land system used with Air Force’s RQ-4 Global Hawk UAVs weighs too much.) Mr. Young also didn’t say that the Air Force has made great inroads in reducing Predator accidents by forward deploying operators in theater to handle MQ-1 landings and recovery. One of the leading contributors of Predator mishaps during landing was the transmit delay when Predator operators stationed stateside sent command signals to the aircraft thousands of miles away. This is alleviated by having operators in theater.

Air Force magazine’s “Daily Report“, April 30th, 2009.

The link in that quote goes to this AFA blurb, which in turn links to this DoD Buzz post, which contains a lot of background data.  But for me, the bolded comments are the interesting part.

The original UCAV concept of operations was to have a fleet of UCAVs in the AOR being controlled from a CONUS-located facility.  But as the operators found out, signal delay was causing a fair number of accidents and hull losses, which can get to be expensive.  So the problem was solved by moving the operators in-theatre as well.  This is all fine and good for strike UCAVs operating within a permissive environment, where control of the air is unchallenged.  But it would not be a tactically sound arrangement for a more challenging environment.  In other words, if those UCAVs had to fly and fight against enemy aircraft, any signal loss or delay would be disastruous.

Certainly an air dominance UCAV would have a manoeuvring edge against a traditional, human-pilot-on-board opponent; it would not be restricted to the G-force limitations of the human body.  But would that be sufficient to overcome moments where the UCAV might be temporarily “adrift”, flying on the last commanded heading and course before the digital umbilical was interrupted?  Would we not also have to co-locate the UCAVs and their pilots within the same theatre?  That might solve the signal delay issue, but it introduces much more risk.  Because pilots in trailers parked at an air base are much more vulnerable than pilots astride their afterburning steeds.  Such an arrangement would present a premium (indeed, unavoidable) target to the opposition.

The solution to that, then, would be to incorporate some kind of auto-fight feature, which would allow the UCAV to continue to prosecute its airborne target in the absence of positive signal from the control trailer.  That way UCAVs could return to their original concept of operations, with secure stateside controllers, and only the UCAVs themselves, the maintainers, armorers (plus logisticians who supply them) close to the enemy and potentially in harm’s way.  But then you run into the complexities of how to teach a UCAV AI how to tell friend from foe (especially in ad-hoc “coalitions of the willing”, where the friend may not share our IFF gear, or worse, share the same type of gear as the enemy), how to identify the posture of its opponent, likely counter-moves, and how to best optimise its own flight characteristics to get the critical shot.

So my sense is that air-to-air UCAVs are still a long way off, until satcom speeds can be significantly boosted, or ways found to mitigate the risk of in-theatre pilots / control trailers / etc aggregated together into a large, tempting target.

You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.

Comments are closed.