Air Force Tilts Toward Unmanned Aircraft

May 2011
By Robert K. Ackerman, SIGNAL Magazine
E-mail About the Author


A U.S. Air Force remotely piloted aircraft (RPA) flies over the desert in this artist’s concept. Increasingly capable RPAs will constitute a significant number of aircraft in the future Air Force arsenal.

Seat of the pants may give way to fit of the chip.

Future aircraft that break into several pieces mid-flight may represent a technological advantage rather than a catastrophic incident. The U.S. Air Force is looking to develop unmanned aircraft that introduce a new set of capabilities not available with humans in the cockpit. Among these many future possibilities are transformer-type vehicles that split into separate flying segments and then reattach when their mission is completed.

Some of these advances lie many years in the future, but Air Force researchers are building a technological road map to reach these destinations. Many innovations, such as signal processing and autonomous software, will be needed to attain these new capabilities. Many other innovations, such as advanced sensors, will be needed to properly exploit these capabilities.

While the efficacy and developmental timetable are uncertain, one trend is proceeding apace. The Air Force is moving pilots out of the cockpit and into the control room. Unmanned aircraft are playing a growing role in the aerial warfighting force, and that growth likely will not abate in the foreseeable future.

That trend already is evident with today’s pilot trainees. The Air Force now is training more pilots to fly what it calls remotely piloted aircraft, or RPAs, than it does for conventional aircraft, says Dr. Mark T. Maybury, chief scientist of the U.S. Air Force. This points out the significant role the service expects RPAs to play in the near future.

Over the next 20 years, RPAs will be carrying out an increasing number of Air Force missions, Maybury offers. They already have been conducting intelligence, surveillance and reconnaissance (ISR) successfully, and new mission areas such as search and rescue are beckoning. Larger craft may engage in resupply, he suggests. “There is a bright future for RPAs, and one of the challenges we face is advancing ideas to enable them.”

Unmanned aircraft will not assume all air combat roles, Maybury continues. Yet, they will be performing more diverse missions than currently capable. The shift will occur along the lines of missions that have been characterized as “dull, dirty and dangerous.” These are the situations out of which the Air Force would like to remove humans, so it is pursuing its RPA research in those directions. Dull missions would be those that are repetitive and do not require much flying skill; dangerous missions would threaten the lives of a pilot or crew; and dirty jobs might be those that combine dullness and danger.

RPAs already have assumed a lion’s share of ISR missions. These aircraft are able to keep station for longer periods of time over areas of interest than their crewed counterparts. Some craft can linger for days and even weeks, Maybury observes, adding that future ISR RPAs may provide persistent surveillance for months at a time.

The Air Force expects that evolving machine capabilities will render human involvement “the weakest component” in combat aircraft over the next 20 years, he says. Foremost among these will be an aircraft’s ability to perceive. Sensor systems aboard existing aircraft already can out-perform their human sensory counterparts. Today’s RPAs have cross-modal sensors such as electro-optical cameras, infrared sensors and signals intelligence packages that allow them to hear and see aspects of their airborne environments.

Future aircraft will be able to touch, smell and maybe even taste their environments, Maybury predicts. Ongoing work is producing acoustical detection that can determine both range and location of a shot fired at the aircraft, for example. Maybury characterizes this as a “super ability” that humans do not have.

The precision-guided weapons aboard existing aircraft are relatively primitive compared to the improved precision of future airborne weapon systems. Maybury predicts that the more sophisticated future weapons will allow operators to control the effects of their use.

And, some of the usefulness of these advanced precision weapons will emerge from their being borne aloft by RPAs. “Because [RPAs] give you loiter time, they provide you tactical patience,” Maybury explains. “Because they provide you with the ability to have not just the pilot and the sensor operator in the loop, but also the judge advocate or military commanders, they provide a much more inclusive decision-making process.

“Even though we’re automating the systems, in some respects we’ll actually make them more human,” he contends.

These unmanned vehicles will require greater autonomy in their operation. Maybury notes that activities such as reasoning and response are important for timely operations. For example, detecting and responding to threats will be a vital area for automation. New technologies will be developed to augment or replace specific human action. With advanced autonomy, humans will become more “on the loop than in the loop,” he states.

To reason, understand and act effectively, RPAs will need to be able to process information rapidly. Maybury notes that one solution may be cooperative autonomous vehicles, and that trend already is appearing. For example, multiple distributed RPAs might be able to perform more effective search and rescue operations over a broad area.

The Air Force will want its RPAs to have greater autonomy when it has “a need for speed,” Maybury allows. For example, the unmanned vehicle would need to be able to respond quickly to changes in airflow over its control surfaces. These machines also would need better detection capabilities for both targets and threats.

But one of the biggest hurdles facing increased autonomy in RPAs is the ability to have certifiable trust in autonomous systems. Maybury maintains that achieving this task will challenge the scientific community to prove that a system will operate as expected under a wide range of conditions, both environmental and operational.

Improved autonomy does not mean that the Air Force will be launching independently operating aircraft. Maybury emphasizes that, for the foreseeable future, all aircraft likely will be piloted remotely. An increasing number of functions will be taken over by technology, but the human will remain in the loop. That human will be necessary to ensure or complement that certifiable trust, he notes.

He cites as an analogy how automobiles increasingly have assumed many functions from drivers. The operator still steers the vehicle and pushes the gas pedal and the brake, but signal processors convert those actions into enhanced performance by regulating fuel mixtures or controlling the application of the vehicle’s brakes. Future combat aircraft, whether piloted directly or remotely, will be more automated in function, navigation and operation.


Future Air Force RPAs will comprise fractionated aircraft that perform different duties while networked as they swarm throughout a battlespace.

Air Force RPAs will have a range of new capabilities driven by opportunity and necessity. These aircraft are likely to have more onboard processing, which will help mitigate one of the major challenges facing greater use of unmanned aircraft—communications bandwidth. Even a highly autonomous RPA will require downlinks to relay its sensor data, and broader-area sensors that are collecting data-rich imagery pose the threat of jamming the communications pipes.

So, the RPAs will have to be more intelligent. They will be responsible for sifting through possible imagery to separate the digital wheat from the chaff. In some cases, this will occur before actual collection. Some RPAs will scout out and identify anomalies, possibly through cross-sensor fusion, to determine which objects are valid targets of interest.

And, the aircraft themselves will be smarter aviators. Maybury foresees greater cooperation across multiple vehicles and between manned and unmanned aircraft. Some RPAs would be expendable vehicles that would perform active countermeasures or fly into hazardous airspace to identify risks that they would report back directly to crewed aircraft.

“So, you will see a continual evolution—in some cases, revolution—of the actual [unmanned] aircraft themselves,” Maybury declares. “The sensors that go on that aircraft will be increasingly sophisticated, intelligent, distributed and cooperative.

“We will see a revolution in the ability for systems to perform tasks quickly, accurately and cooperatively,” he concludes.

Cooperative activity among manned and unmanned aircraft only partially describes a future RPA environment. Air Force researchers are pursuing the development of fractionated aircraft that would break up into smaller units that would carry out their missions, and then reassemble into their original whole—“decomposable and then recomposable,” as Maybury describes them.

These fractionated aircraft would be able to operate autonomously or in concert with one another. Early versions might consist of a flying autonomous vehicle that would be dropped by a carrier aircraft that it would rejoin after completing its mission. Or, that deployable vehicle might be an expendable craft.

The next step might be an RPA that is built of different sensor units before it even takes off. “Just like Lego [blocks],” these air vehicles would be customized for their particular mission prior to flight, Maybury offers.

Beyond that may lie RPAs that break up in flight into a variety of smaller vehicles. The mini-RPAs could be equipped with different sensors that fuse data while they are in their independent flights. Then they would re-compose their original vehicle to return to base. Maybury explains that the original RPA could remain intact as a single unit if the target adversary was concentrated, or it could decompose if the mission need were more distributed. This could apply to intelligence gathering or even kinetic weapon attack.

Maybury describes this as a natural progression in the long move toward distributed composable systems. The concept already is in use today in space. Plug-and-play satellites have helped engineers improve successive generations of satellites by changing individual components before launch. This concept has been in use since the adoption of common satellite buses that are equipped with tailored payloads.

Fractionating RPAs provides greater redundancy through distribution of capabilities, he continues. Low-cost expendables can be produced in large volume to make an airborne platform more flexible in mission capabilities. And, if the fractionating RPAs are very intelligent, they could even compose themselves in the battlespace as mission needs dictate. This could involve changing configurations or sensor selections, for example.

A critical enabler for fractionated RPAs is robust communications, Maybury points out. Distributing small vehicles in a highly contested environment mandates that the craft are able to maintain communications links. Their communications suites would need to include low probability of detection, low probability of intercept, jam resistance, frequency agile radio frequency, laser links and burst communications as possible remedies.

“You don’t get fractionation of distribution for free,” Maybury states.

Design is another critical enabler. Fractionated RPAs must be designed to be inexpensive if they are to be deployed in large numbers. Planners must decide which capabilities, particularly data processing and storage, should be onboard and which should be on the ground or on another platform. And, the laws of physics promise that signal latency must be addressed.

Validating and verifying the autonomy is another obstacle. Planners must determine under which conditions an autonomous aircraft will survive. If it is not an all-weather RPA, then it must be able to adjust to conditions to carry out its mission, particularly if it is an ISR mission.


Enjoyed this article? SUBSCRIBE NOW to keep the content flowing.