Drone Swarm Technology: Advancements And Implications – Researchers at [Artist’s Concept] have developed a reinforcement learning approach called Hierarchical Reinforcement Learning that will allow fleets of unmanned aerial and ground vehicles to optimally perform various missions while minimizing the performance of uncertainty on the battlefield. (Photo Credit: Shutterstock) SEE ORIGINAL
ADELPHI, Md. — Researchers have developed a reinforcement learning approach that will allow fleets of unmanned aerial and ground vehicles to optimally perform various missions while minimizing performance uncertainty.
Drone Swarm Technology: Advancements And Implications
Swarming is a method of operations in which several autonomous systems act as a unified unit by actively coordinating their actions.
Cometh The Swarm — Examining The Next Evolution In Drone Warfare
Researchers say future multi-domain battles will require fleets of dynamically coupled, coordinated heterogeneous mobile platforms to outmatch enemy capabilities and threats targeting US forces. .
A small unmanned Clearpath Husky robot, used by ARL researchers to develop a new technique to quickly teach robots new travel behaviors with minimal human supervision. (Photo Credit: U.S. ) SEE ORIGINAL
Looking at the proliferation of technology to enable time-consuming or dangerous tasks, said Dr. Jemin George of the U.S. Research Laboratory. Combat Capabilities Development Command.
“Finding optimal guidance policies for these swarming vehicles in real time is a critical requirement for improving the tactical situational awareness of warfighters, allowing the US to dominate a controversial environment,” said George.
China, U.s. Test Intelligent Drone Swarms In Race For Military Ai Dominance
Reinforcement learning provides a way to optimally control uncertain agents to achieve multi-objective goals when an accurate model for the agent is not available; however, existing reinforcement learning methods can only be used in a centralized way, which requires the aggregation of the state information of the entire population by a central learner. This dramatically increases computational complexity and communication requirements, resulting in unreasonable learning time, George said.
To solve this issue, in collaboration with Prof. Aranya Chakrabortty from North Carolina State University and Prof. He Bai from Oklahoma State University, George is doing a research effort to solve the large-scale, multi-agent reinforcement learning problem. This effort is funded through the Director’s Research Award for the External Collaborative Initiative, a laboratory program to stimulate and support new and innovative research with external partners.
The main goal of this effort is to develop a theoretical foundation for data-driven optimal control for large-scale network networks, where control actions will be performed based on low-dimensional data. in measurement rather than dynamic models.
The current method is called Hierarchical Reinforcement Learning, or HRL, and it breaks down the global goal of control into multiple hierarchies – that is, many small levels of microscopic group control, and one broad-level control of macroscopic level.
Us Army To Test Largest Ever Drone Swarm
“Each hierarchy has its own learning loop with individual local and global reward functions,” George said. “We were able to reduce the learning time by running these learning loops in parallel.”
The researchers envision a hierarchical control for ground vehicle and air vehicle coordination. (Photo Credit: U.S. graphic) SEE ORIGINAL
According to George, the swarm’s online reinforcement learning control is based on solving a large algebraic matrix Riccati equation using the system, or swarm, input-output data.
The researchers’ initial approach for solving this large matrix Riccati equation was to divide the population into several smaller groups and implement group-level local reinforcement learning in parallel while implementing a global reinforcement learning of a smaller dimensional compressed state from each group.
What Is A Drone Swarm: A Comprehensive Overview
Their current HRL scheme uses a decoupled mechanism that allows the team to hierarchically approximate a solution to a large-scale matrix equation by first solving a local reinforcement learning problem and then synthesizing global control from local controllers ( by solving a least squares problem) to run a global reinforcement learning on the aggregated state. This further reduces the learning time.
Experiments show that compared to a centralized method, HRL is able to reduce the learning time by 80% while limiting the loss of optimality to 5%.
“Our current efforts at HRL will allow us to develop control policies for fleets of unmanned aerial and ground vehicles so that they can perform optimally in a variety of mission sets despite the individual dynamics for swarming agents is unknown,” George said.
George stated that he is confident that this research will have an impact on the future battlefield, and is made possible by the new collaboration that is taking place.
Army Advances Learning Capabilities Of Drone Swarms
“The core purpose of ARL’s science and technology community is to create and leverage scientific knowledge for innovation overmatch,” George said. “By engaging with external research through ECI and other cooperative mechanisms, we hope to conduct disruptive foundational research that will drive modernization while serving as a key collaborative link with the global community of science.”
The team is currently working to further improve their HRL control scheme by considering optimal grouping of swarm agents to reduce computation and communication complexity while limiting the optimality gap.
They also investigate the use of deep recurrent neural networks to identify and predict the best grouping patterns and the application of developed methods for optimal coordination of autonomous air and ground vehicles. in Multi-Domain Operations in dense urban terrain.
George, along with ECI colleagues, recently organized and chaired an invited virtual session on Multi-Agent Reinforcement Learning at the 2020 American Control Conference, where they presented their research findings.
Are Drone Swarms The Future Of Aerial Warfare?
. As a corporate research laboratory, ARL discovers, innovates and transfers science and technology to ensure dominant strategic power on earth. Through collaboration with the command’s core technical capabilities, CCDC leads the discovery, development and delivery of technology-based capabilities needed to make Soldiers more lethal to win the nation’s wars and return home safely. . CCDC is a major subordinate command of Drone swarm technology—the ability of drones to make decisions based on shared information—has the potential to change the dynamics of conflict. And we are getting closer to seeing this potential unleashed. In fact, swarms have many applications in almost every area of national and homeland security. Flocks of drones can search the ocean for enemy submarines. Drones can be dispersed over multiple areas to detect and eliminate enemy surface-to-air missiles and other air defenses. Drone swarms could serve as a new missile defense, blocking incoming hypersonic missiles. On the homeland security front, security forces equipped with chemical, biological, radiological, and nuclear (CBRN) detectors, facial recognition, anti-drone weapons, and other capabilities offer defense against various threats.
But while drone swarms represent a major technological advance, unlocking their full potential will require developing capabilities centered around four key areas: swarm size, customization, diversity, and hardening.
In general, the more drones in a swarm, the more capable the swarm is. Larger underwater fleets can cover greater distances in search of enemy submarines or surface ships. Larger herds are more resistant to certain defenses. The loss of a dozen drones will impair the capabilities of a twenty-drone fleet, but not significantly in a thousand-drone fleet.
Media reports indicate that China has successfully tested a fleet of a thousand drones. And China seems interested in the fleet’s capability as a means of attacking US aircraft carriers. Although Intel has deployed a fleet of 1,218 drones, it does not appear to be a true drone fleet, relying on programmed behaviors rather than inter-drone communication.
Seeing Through Forest With Drone Swarms
There is little reason to believe that herd size will not continue to grow significantly. Building a large crowd primarily requires the ability to handle large amounts of information. More drones means more inputs that can affect the behavior and decisions of the herd. And on a basic level, more drones means a greater risk of one drone crashing into another.
Of course, the importance of crowd size will depend on the mission. Stealthier missions against slower targets don’t require thousands of drones. And too many drones can be devastating, drawing unnecessary attention from defenders. But large-scale attacks on enemy bases and other difficult targets may require that. An attack on a hard target means a greater expectation of defeat and a greater need for a strong offensive.
The future drone swarm will not necessarily consist of the same type and size of drones, but will include large and small drones with different payloads. Joining a diverse set of drones creates a whole that is more capable than the individual parts. A drone swarm can even operate across an entire domain, with undersea and surface drones or ground and aerial drones coordinating their actions.
Current drone swarms are primarily small, identical, sensor drones, but simple multi-domain swarms have been developed. One such fleet includes a flying drone partnered with a walking drone. The aerial drone maps the immediate area and the ground-based drone uses that information to plan its actions. Another experiment showed a fleet of five different-sized drones on the ground, each with different sensors and different functions. The five drones work together to perform a basic human rescue mission, transporting a dummy to a different location.
Emerging Technology And Risk Analysis: Unmanned Aerial Systems Intelligent Swarm Technology
Drones within the fleet can serve different roles based on their different capabilities. Attack drones conduct strikes against targets, while sensor drones collect information about the environment to inform other drones, and communication drones ensure the integrity of inter-swarm communication. .
Small sensor drones can provide reconnaissance for larger attack drones, collect information on enemy targets and relay it to attack drones to conduct strikes. Even the drones specifically tasked with conducting attacks can be different. A drone swarm can involve attacking drones on
Technology advancements in education, medicine and technology advancements, science and technology advancements, united states technology advancements, space technology advancements, advancements in transportation technology, military technology advancements, swarm technology, technology advancements in healthcare, drone swarm, technology advancements in business, drone swarm technology