By BROOKS TIGNER
BRUSSELS – Just as policymakers and the public are starting to grasp the idea that fleets of identical drones can be programmed to work together, i.e., to “swarm”, – like a flock of intelligent birds, research is taking the concept to a higher – and more complex – level with applications relevant for military and civil security users.
The idea is to enable the swarming not only of different drone models but intelligent platforms of all kinds, whether in the air or on the ground.
Indeed, the European Defence Agency and national defence technology experts are considering how carry forward the results of a recently completed research project funded by the EU. The project entailed a preliminary investigation of how to swarm heterogeneous sensor platforms.
The short, 15-month project, known as EuroSWARM, identified two central problems that future research must address, namely: how to achieve enough autonomous computing power for networked sensors and how to ensure the latter can accurately identify threats, and particularly those arising from so-called behavioural anomalies. The project ended its work in February 2018 and was carried out by defence and university research centres in Greece, France, Sweden and the UK.
“Behavioural monitoring requires enormous computational complexity,” an EDA official recently told SECURITY EUROPE. “Swarmed network currently can’t do that without [computational] support from a control centre.”
Several new algorithms emerged from EuroSWARM that enable operators to dynamically reconfigure a swarm’s sensor fusion, cooperative guidance and other tasks. These were demonstrated at facilities provided by the UK’s Cranfield University, which led the project.
EuroSWARM did not develop specific algorithms for automatic target identification, however, “because that requires more work. Yet they developed several new algorithms, including swarmed task allocation and path planning,” said the official.
The project integrated small, cheap, off-the-shelf drones with a trio of unmanned ground vehicles and some static sensor devices. Due to budgetary restraints – EuroSWARM’s budget was a modest EUR 490,000 – the research team had to use simulation to test how the static and mobile platforms worked together in their scenario-based exercise. The scenario entailed anomaly detection and tracking to protect a military camp.
The computing power required to detect anomalies exceeded the project’s capabilities, but the prototyped system proved it could track anomalies once they were identified.
That functional weakness points to the focus of future research: to enable swarmed systems of mixed models and makes of ground vehicles, air platforms and remote stations pulse as one for purposes of detection, identification and tracking of threats.
“The swarm successes you hear about elsewhere such as in the United States and its Perdix [micro aerial surveillance drone] experiments have involved homogenous fleets,” said the official, “Heterogeneous swarming is much harder.”
But heterogeneous swarming could also have many positive applications for civil security, from fighting fires to search-and-rescue to counter-terrorist operations. Terminator fears should not prevent researchers from at least exploring the concept. Indeed, this is a technology whose research could be financed either by the EU’s civil research programmes or via the “research window” of its forthcoming European Defence Fund, which will disperse some EUR 500 million a year, starting in 2021.
One thing seems certain, however. Allowing swarmed technology to seep unfiltered into society merely as a market “freedom” could be dangerous. Policymakers need to follow this kind of research to anticipate the laws necessary to ensure it doesn’t run out of control to the detriment of ordinary citizens.