Drone swarms are already a technical reality but how can we control them safely and efficiently, so they operate the way we want, as a collective, as groups, and at the individual level? RISE researchers turned to the world of gaming to find the answers.
In the future, drone swarms could play a crucial role in critical societal operations, such as responding to natural disasters or supporting rescue missions in complex environments. But for this to become reality, more advanced ways of managing and deploying autonomous drone swarms are needed, without losing control.
Oscar Bjurling, a researcher at RISE, became interested in this challenge and began studying human-swarm interaction during his doctoral project. He realized that an entire industry has already grappled with similar problems for decades: in real-time strategy games, hundreds or thousands of agents are managed simultaneously, often under intense pressure.
The project led Oscar and his colleagues at RISE to develop an interface for a drone simulator, for research purposes and inspired by game design patterns, where complex missions such as wildfire suppression or search-and-rescue operations can be simulated.
Game interfaces inspired design solutions
In the project, Oscar took a closer look at the games Ashes of the Singularity, Total War: Warhammer II, and Drone Swarm. He selected them because they differ in scale and control methods: from managing individual units to handling battalions or swarms of up to 32,000 drones. By analyzing their interfaces, the researchers identified design patterns that could be transferred to real-world ground control stations.
The goal of the project was to identify and understand the operational challenges faced by swarm operators and to develop design solutions at both system and interface levels that enhance collaboration between humans and drone systems. The challenge lies in controlling groups of drones, or an entire collective, while maintaining oversight at the individual level. Today, significant progress has been made.
“Drones are now autonomous enough that we shouldn’t need to control them in detail. Instead, we can use systems with predefined tactics and algorithms that act as building blocks,” says Oscar Bjurling.
The human role crucial for the swarm
He explains that the goal is delegation: instead of micromanaging each drone, the operator should be able to give high-level instructions that the system then executes independently. This implies a hierarchical system where humans make the key decisions while the swarm handles execution.
Technically, we may be close but today’s drone ground control systems are still limited to controlling one unit at a time. In addition, each drone currently requires a safety pilot who maintains constant visual contact with it. This makes practical swarm deployment impossible for now, and important challenges remain before we can see drone swarms in the air.
These challenges include adapting solutions to specific applications, understanding when and how humans need to intervene, establishing regulations for autonomous systems operating beyond visual line of sight, and achieving viable business models for costly systems.
The importance of testing and verification
Setting legal barriers aside, Oscar believes the technology could be operational within a few years. However, he urges patience:
“When systems are autonomous, it’s extremely important that they are thoroughly tested and verified. This will take time, and it needs to.”
Within ten years, he is convinced that drone swarms will be useful in applications such as wildfire fighting and search-and-rescue operations.
RISE is now working to develop next-generation drone swarms to support Swedish industry. By combining research, test environments, and broad engineering expertise, RISE contributes to creating safe, reliable, and scalable solutions, enabling technological breakthroughs while ensuring that the technology can be used responsibly and effectively in critical applications.
