The swarm intellect can reduce the effectiveness of most modern weapons
Written by Vasily Burenok; Originally appeared at VPK, translated by AlexD exclusively for SouthFront
One of the promising directions of development of military hardware considered is the use of remote-controlled devices. The talk is not only about Unmanned Aerial Vehicles (UAVs), but also about ground, marine as well as space drones. Nevertheless the expansion of their use will create problems.
Each robotic system is not only an execution object, equipped with reconnaissance equipment and data transmission, but in certain cases it is armed as well. It is a complex system, which includes the machine control, guidance system, positioning, communications, data processing as well as an operator, controlling the drone.
Collective Intellect
Unfortunately, the fascination of the development of such devices led to the fact that each one of them is unique, consequently, as many drones or ground robots we see, there are as many different control systems developed, and essentially do not interact with each other. Obviously this cannot go on indefinitely.
One of the directions for solving the problems of controlling robotised vehicles was the introduction of the so-called swarm intelligence. Gerardo Beni and Wang Jing introduced the definition of the term in 1989, not defined until then. Swarm intelligence is described as the collective behaviour of decentralised self-organising systems.
The technology of group control (swarm) of robotic tools in a simplified explanation must ensure the implementation of principles of joint actions of a group of insects, fish or fowl, which can share information and carry out the collective task, using the collective intellect, complementing each other and not interfering each other in the process.
It is believed that a single human operator will control this drone swarm. While such tasks as the construction of the military order, depending on developing situation and the nature of the task manoeuvres in the pack, and a few others will be resolved without human intervention, they will fully take on the swarm intellect.
The full-scale implementation of swarm intellect principles in the robotic devices for military use will depreciate many of the expensive systems in service that are found in the arsenals of Russia as well as other countries. For example, the cost of current aviation complex of tactical fourth generation frontline aviation amounts to tens of millions of dollars, of fifth generation, it is even higher. The use of anti-aircraft missiles of the S-300, S-400, S-500 systems for destruction of such an aircraft are justified, since the cost to run the anti-aircraft missile system (AAMS) combat missions are significantly lower. Therefore the actions of modern aviation in the air defence zone of the enemy, breaking through the air defence, are a too expensive. Accordingly, the aforementioned AAMS is very effective.
But everything will change drastically if on the battlefield hundreds of drones appear, so using the already expensive AAMS will be like to trying to kill a fly with an iPhone. And if the drones will be capable of exchanging information on the positions and actions of the air defences, carrying out air manoeuvres, interfering with radar systems, striking air defence systems and covering up their equipment, then to talk about an effective defence system on the basis of existing AAMSs becomes meaningless.
The range of problems to be solved by the swarm can be arbitrarily wide. For example, the search and blocking of ballistic missiles launchers.
In a conventional war they can become a tool not only in the defeat of the enemy, but also as psychological pressure and demoralisation tools. The report by Stanford and New York Universities “Living under Drones” in September 2012, described the following result of the use of American drones in North Waziristan. Around sixty drones flew over the local villages practically around the clock, able at any moment to strike at any object. It was believed that terrorists were in that location. But in reality no one knew how people and structures fell into which category to be destroyed. In the end, the livelihood of a vast territory was completely paralysed by fear because of death constantly hanging in the air.
Tremendous opportunities for strikes on any given point on the globe open up with the placement of swarms of drones on aircraft carriers. American experts have long understood this, and so the USA very successfully is developing the specialised shipboard drone X-47V Pegasus. There is no doubt that drones based on aircraft carriers will be equipped with swarm intellect.
American specialists went even further in the realisation of the Gremlins project on the orders of the Defense Advanced Research Projects Agency of the USA (DARPA). According to the developers the swarm drones will be based on aircraft carriers having the possibility of massive take-offs and landings.
“Locust” in a Container
One of the key moments in the development of the swarm technology is the creation of the information sharing system between separate drones. The University Johns Hopkins Applied Physics Laboratory announced in 2012 the development of the mobile peer-to-peer network, which allowed the drones to exchange information during flight.
The same year, specialists at Boeing demonstrated for the first time the systems operation of advanced drones control, using the swarm technology. During the tests two Scan Eagle drones by Boeing and Procerus Unicorn made a joint flight, independently exchanging data, necessary for the completion of the task. It was to find a specific site on the ground. So the drones jointly scanned the terrain, drew the flight map and sent data to the control centre.
A single operator using a laptop and radio operates the drone flights that operate autonomously. Incidentally, the transmission of commands was done without the use of staff in ground stations, which demonstrated the simplicity of operation and the possibility of reducing costs of “conducting” a swarm of drones. No wonder the specialists at Boeing consider that their experiments can become important milestones in the development of drones.
It is expected that in the future swarm technology will facilitate the operation of large numbers of robots for the operators-pilots; they will only need to set goals and rely on the capacities of the swarm intellect. Another innovation that is expected from this technology is the possibility of plugging in the information channels of the swarm by any users in the combat field.
On the orders of the Office of Naval Research of the US Department of Defence the Low Cost UAV Swarm Technology (LOCUST) project is developed, which sounds similar to low cost. It is supposed to create drones for land, air and naval bases, which are stored in containers and are released from it. The system is designed for the detection of airborne objects, including low-flying and low-visibility targets at an altitude of 10 metres and a range of 100 to 200 kilometres respectively with high accuracy (up to 0.2 metres per second) in determining the speed of their flight.
Thus, the above described swarm systems, created by the USA, will be completely different in comparison with the existing patterns of warfare with enemy air defences and organisation of their own defences. They will negate the capabilities of existing air defence systems and will help create a more effective and less expensive aerial reconnaissance.
The control of swarm technology is applied not only in UAVs. NASA’s engineering allowed the Americans to create marine swarm robots. In 2014 on the river James, near the entrance of Fort Eustis (State of Virginia) the U.S. Navy conducted an exercise involving 13 robotised boats that acted as a single group. The special swarm architecture for control CARACaS was used. It involves optical-electronic and radar reconnaissance, ship identification, enemy recognition and situation analysis (Contact Detection and Analysis System), as well as control of movement of the robot. It is planned that the above-mentioned will be executed in the form of options installed on any surface ship.
The swarm control algorithms were initially developed multipurpose to allow their use for applications on ground, air and, of course, space technology. So far the main task of a swarm of robotic boats is the protection of ships, ports, oil platforms, other marine and coastal structures.
It is assumed that on robot-ships it can be installed remotely or automatically controlled machine guns of 12.7mm calibre and other weapons. Thus a swarm without human intervention can block and out of necessity destroy terrorists, pirates, small and big warships as well as similar groups of enemy robots. But the negative experience in the applications of American autonomous robots in Afghanistan raise doubts to the effects that this is done without authorisation from the operator. So, we will repeat, questions remain that need to be addressed.
The Denial of Denials
Equipping boats and ships with this kind of serious weaponry will allow them to fulfill more complex tasks: blockading channels, laying barrier minefields, war on sea and ocean passages. One cannot doubt that this kind of technologies is developing.
As mentioned above the American space agency NASA, would have to say that the technology of swarm control has long been regarded as promising for space. For example, instead of single large spacecraft for solving problems of remote sensing of the Earth, the creation of a group of small spacecraft is found to be effective and is already being implemented, equipped with apparatus for various purposes, for example visible, infrared, radar spectrum. The small spacecraft grouping must not only independently make decisions but, if necessary, negotiate and have the flexibility to form coalitions or teams of satellites for various purposes, including new ones, originally not strictly specified in advance. Thus the signal, suddenly triggering a change of plans throughout the group, can come either from Earth or any of these satellites, for example from the first to detect the dangerous new object. In this sense the entire satellite group manages collectively with the participation of each of them, because each members of such a team can send an alarm and notify the other satellites about creating of new team.
The work of such systems, where a complex object consists of a group of autonomous functioning, but constantly interacting parts, requires completely new approaches, methods and tools for building these operating systems. Obviously NASA has achieved significant progress in this direction, and it permitted the diffusion of similar technologies to control other objects.
It should be noted that, as in the cases above, the use of swarm control in space largely eliminates the known methods of neutralising orbital groups.
Of course, one cannot presently talk about swarm systems as a perfect weapon for modern war. A number of very important and complex tasks are not solved. First is the stability of the control system and data exchange in a strong radio-electronic fight. The use of electronic warfare can block the functioning of the swarm and reduce its effectiveness to naught. It is also well known how intensively photonics are developing, which may be a way to parry efforts of electronic warfare.
The next problem is the development of adequate algorithms (on the human level of intellect) for processing the incoming information from elements of the swarm and adoption of appropriate optimal solutions. This is particularly important in the case of application of means of destruction. It is clear that people can make mistakes, but can the swarm intellect make a decision in a difficult situation, will not carry out fatal mistakes, which will lead to not only local, but bigger disasters?
It is assumed that information from different types of sensors will be used. This is a difficult task, which in addition to the complexity of the processing algorithms will come up against problems of absolute necessity of preparing large amounts of intelligence, elaboration and adoption of optimal decisions in real-time onboard the UAVs. Consequently, computing power, energy for their support will have to meet very stringent requirements. And this, as stated by American specialists, it is difficult to implement in a simple, small and cheap drone.
But sooner or later all these problems will be solved. We must not forget this when creating increasingly sophisticated and costly weapons on the basis of know principles and designs. Even outstanding examples from the point of view of the present situation may be useless in encounters with new technologies.
Vasily Burenok, President of the Russian Academy of Rocket and Artillery Sciences, Doctor of Technical Sciences