Overflights of nuclear power stations, irruptions on runways of airports and other untimely falls, but also targeted attack, like that perpetuated in Yemen last January and, henceforth, large-scale military attack, such as the one conducted on Saturday, September 14 in Saudi Arabia against oil installations … While the all-round development of drones has opened up many opportunities, these machines also constitute, in the civilian and military world, a diffuse, permanent threat, and – in financial terms – within reach of any stock exchange .
The counterattack is a puzzle. How to detect such small devices moving at such a high speed? How to distinguish in a few seconds a friendly UAV from a malicious device? And after identifying it, what technologies can be used to neutralize the threat? It is to answer these questions from a technological point of view that the National Research Agency (ANR) launched a call for projects in 2014, on behalf of the General Secretariat for Defense and National Security ( SGDSN). An initiative that has helped mature technologies, placing France today in a good position in the fight against drones.
Detect, identify, neutralize
“Following the first flights of illicit UAVs on sensitive sites, between the end of 2014 and the beginning of 2015, France is one of the most responsive countries,” said Stéphane Morelli, Institutional Relations Manager at Azur Drones. at L’Usine Nouvelle, many of these projects have reached maturity and some are already on the market. ” A rise in power of a very specific know-how, which one owes to a big group such as Thales, but also to the manufacturers and operators of drones, and to dozens of specialized companies like CS.
The immutable triptych to counter an attack: detect, identify, neutralize. In France, the most successful implementation in this area on a critical and large site is undoubtedly at the Roissy-Charles-de-Gaulle airport. Since the beginning of May, the tenth largest airport in the world has been equipped with a real anti-drone arsenal, called Hologarde. A system – at a cost of about 5 million euros – which will have to protect a total surface of 6,000 hectares by the end of 2019, compared to 3,200 hectares of Paris-CDG.
Paris Airport (ADP), associated with the Directorate of Air Navigation Services (DSNA), has been surrounded by a team consisting of Thales, the Gard SME Exavision and the start-up Toulouse InnovATM. If this device proves in the coming months its effectiveness, it could perhaps make emulators and generalize on several large sensitive sites.
To ensure the detection signal jammer and identification steps, the group of companies has developed a goniometer, which detects all frequencies transmitted by or to drones, compares them to a data library and provides classification information. What is added a holographic radar, able to detect very small, very slow targets, and this at great distances (7 km), but also to sort out the nature of the targets (drones, birds, vehicles …). Other equipment used: military surveillance cameras with a range of 3 to 4 km. The whole thing is headed by a software which analyzes the trajectories, evaluates the mass of the drone, and which, in fine, warns the chain of command, limiting to the maximum the false alarms.
When artificial intelligence gets involved
“The detection systems are now quite efficient for the most common UAVs,” said Stéphane Morelli from L’Usine Nouvelle, “but we must recognize that the neutralization phase remains problematic.” In the civil first, because it can be assured only by the police. But also because the technologies must be able in a few seconds to neutralize a small device … Drone drones equipped with net and large birds do not seem more relevant today.
Scramblers, portable or placed on mobile turrets, may provide a solution, but may also interfere with all surrounding communications. Direction-finding can emit waves and cut the link between the drone and its base, or even take control. The ADP group has moved closer to the start-up Roboost to develop systems of remote control of the drone. Tests are already in progress. Still, pirate drones could protect themselves … The next step will probably be to automate the response to gain speed and efficiency, which requires the development of specific algorithms or even the use of the artificial intelligence.