Service Academies Swarm Challenge: Expanding the capabilities UAV swarms

launching as many UAVs as quickly as possible and keeping them aloft as long as possible. The team that won both of its matches won the competition.

The three matches saw successively higher scores and more complex gameplay, as the teams gained and learned from their experience:

— Sunday, 23 April: In this opening game, Air Force defeated Army 58-30 in a 20-on-20 battle. Air Force’s 10 fixed-wing UAVs and 10 quad-rotors had 13 air-to-air tags and no air-to-ground tags, while Army’s 15 fixed-wing UAVs and five quad-rotors had six air-to-air tags and one air-to-ground tag. Army took an early lead but Air Force’s ability to quickly amass a larger swarm proved to be the decisive factor.

— Monday, 24 April: This classic match-up saw Navy defeat Army 70-37 in a 20-on-25 battle. Navy’s 11 fixed-wing UAVs and nine quad-rotors had 16 air-to-air tags and five air-to-ground tags, while Army’s 17 fixed-wing UAVs and eight quad-rotors had eight air-to-air tags and three air-to-ground tags. The match was all about the fixed-wing aircraft, with only one successful air-to-air tag by a quad-rotor.

— Tuesday, 25 April: Navy defeated Air Force 86-81 in a hard-fought championship match in which the lead changed four times before the clock ran out. Both teams launched every aircraft the game allowed, fielding the most of all the matches with 60 UAVs aloft at a time—25 on 25, while each team launched and held five additional craft in reserve. Navy’s 20 fixed-wing UAVs and 10 quad-rotors had 19 air-to-air tags and six air-to-ground tags, while Air Force’s 18 fixed-wing UAVs and 12 quad-rotors had 18 air-to-air tags and four air-to-ground tags. The large number of aircraft pushed the experimental networking infrastructure to its limits and made it harder for both teams to send commands and update tactics.

Each team developed multiple tactics as well as an overarching strategy for implementing them during matches. DARPA is currently evaluating the data from this preliminary research for potential applicability to future military missions.

Looking Ahead
The Service Academies Swarm Challenge highlighted many potentially fertile areas of future swarm-related research, Chung said. Topics of particular interest include efficient intra-swarm communications, improved situational awareness among swarm members, and developing human-machine interfaces and other tools that allow swarm commanders to more effectively use swarm capabilities. It also underscored the value of installations such as Camp Roberts, which has the restricted airspace, infrastructure, and experience to host large-scale UAV experiments.

The three Service Academies are poised to extend the research that supported the Challenge. They are keeping all the UAVs, software, and other infrastructure that DARPA provided. And they have created basic curricula and capstone project frameworks for conducting unique, accelerated research and field experiments on swarm tactics. Most important, they now have dozens of students who have worked side-by-side with leading-edge researchers and operators experienced with advancing the state of knowledge of swarm systems.

“There are seeds of genius in these Cadets and Midshipmen and I hope that one day some of them will return to DARPA as program managers,” Chung said. “There’s a bright future for these young military leaders, in that they now—having experienced the cutting edge of technology—can help shape how that technology will be used in the future.”