DARPA’s Game Of Drones

Image: 42levelone.com

For his story “Willie Pete Has No Off Switch,” Alec Medén was a finalist in the Atlantic Council Brent Scowcroft Center on International Security Global Trends 2035 creative contest that called for writers to explore the technologies, trends and themes that will shape the world two decades from now. He is also a past winner of the project’s “Space” war-art challenge with “From A Remove,” judged by bestselling author David Brin, for a story that envisioned the future of space conflict during the last decade of the 21st Century. Alec is majoring in screenwriting and creative writing at Chapman University in Southern California. He can be found on Twitter @alecmeden.

The cloud of drones swirls like a flock of starlings around Seoul skyscrapers. An embattled fireteam of Marines pepper the swarm with small arms fire, but they can’t make a dent in the sheer mass of coordinated metal. Before the machines descend on the team, the Marines barely manage to set off a Non-nuclear Electromagnetic Pulse (NEMP) device, turning the entire swarm into a heap of inert metal and plastic.

In this scene from the blockbuster video game Call of Duty: Advanced Warfare, future technologies greatly desired by modern day military thinkers are tested on the battlefield of imagination. Perhaps in a world where video games tend to resemble modern conflict more and more, gaming stratagems precede tactical realities. Whether or not that presages the future, it certainly seems to be echoed in a new Defense Advanced Research Projects Agency (DARPA) Program, the recently announced initiative aims to create a “real time, networked virtual environment that would support a physics based, robotic swarm tactics game,” where a curated group of players would develop winning strategies for urban conflict by resolving tactical situations in gameplay, which could then be tested and applied to real world situations. To control hundreds of these drones both in simulation and in real life, DARPA is asking for a user friendly, intuitive interface to control dozens if not hundreds of component drones, in addition to creating a “playbook” of premade actionable strategies not unlike what coaches use in football.

How then might this actually move from the screen to reality?

First and most obviously, the concept implies the use of swarmed robots in a battlefield role, with an offensive element. Weaponized quadcopters immediately spring to mind, although one can easily imagine a far more diverse set of unmanned vehicles patrolling the air and ground of a battlefield. Secondly, this program shows a willingness to embrace crowd-sourced solutions not only with technology (as seen in programs like DARPA’s Adaptive Vehicle Make) but also in combat tactics. Obviously there may be meaningful gaps in the difference between the simulated space and the complete chaos of real world combat. Despite that, the results of such a program may be hugely instructive, especially for a military technology so early in its infancy. Swarming is especially conducive to this kind of rapid and unconventional crowd-sourced innovation, as the relative expendability of the components of a swarm allow for more experimentation and less rigidity and caution.

This program is not without precedent. In the medical sector, Phylo is a striking example of the effectiveness of such “gamification” methods, using what was essentially a game to allow players on the Internet to optimize nucleotide sequence alignments for complicated phylogenetic research. Reminiscent of “Smartmobs” in David Brin’s novel Existence, or Ender Wiggins’ control of entire fleets in Ender’s Game, this process uses human brainpower and judgment to solve complex issues. This method is especially applicable in situations where it would be cost prohibitive or even impossible to use an algorithm or other computational means to find the solution. Perhaps this is the beginning of more revolutionary tactics and technology, as nations and nonstate entities use networked human minds incentivized by gamification to obtain results from previously intractable problems.

Indeed, DARPA has also recently opened up a program known as Gamifying the Search for Strategic Surprise, which, true to it’s name, will use a network analogous to online games and social media platforms to connect and motivate individuals all over the world to help brainstorm possible surprises and upsets in the future military climate.

Moving into less grounded and more extrapolative territory, the DARPA algorithm program also begs the question of what risks the employment of drones in this capacity might yield. Skipping over the ever-present criticism of overreliance on technology, the biggest pitfalls might be ruggedization, informational overload and susceptibility to enemy action. It seems hard to imagine an intuitive, highly advanced interface, haptic or otherwise, that could survive conditions in the variety of urban battlefields the military might face – or be developed fast enough. After all, the comparative lag between the functionality of technology in the Land and Nett Warrior programs and civilian smartphone technology shows that on a contested battlefield, technology becomes just as difficult to control as the humans wearing it, if not more so. Just as soldiers are trained, so to is equipment stripped down, built up, or otherwise changed to meet mission requirements. I imagine it would be horrifying to watch a swarm of drones lie inactive and exposed to enemy fire all thanks to some dust in your computer tablet or haptic gloves. That sort of low-tech snafu never seems to happen in video games.

Another issue may be informational overload. Even with a remarkably intuitive control system, any soldier who must control dozens or even hundreds of frontline combat units (as each drone must necessarily be) would experience at least a slight denigration to their other warfighting abilities. Due to the requirement of focus, expertise, maintenance, and battery life, operation of these drones may be delegated to certain individuals, much like how calling in close-air support is delegated to a forward air controller (FAC) who has expertise in this specialty. Perhaps in the future, urban drone operators might become similar to an FAC, focusing on their specific combat capability rather than operating in the same way as the rest of the riflemen in their platoon or crewmen in their vehicle.

As for the dangers of inevitable enemy tampering, any adversary capable of engaging in some kind of electronic or cyber warfare would attempt to divorce these swarms from their controller. From brute force jamming to subtle intrusion, if these unmanned vehicles are controlled directly and remotely, this control system could be interfered with. Perhaps partial autonomy might solve this problem, or maybe this will emerge as another battlefield risk to mitigate.

Regardless of how these issues are resolved, if DARPA receives a workable swarm tactics game, how could any gamer resist taking it for a spin?