From birds to drones. How swarming intelligence is creating a new being
Through history, our relationship with nature has been rather pedagogical. We have not only served ourselves from nature to eat and live, but we have hugely benefited from it by learning about collective behaviours. “Nature is the art of God” says the phrase related to its marvellous design where everything and everyone interact.
From fishes in the deepest ocean to birds in the sky through to bees in the field, it’s easy to identify an incredible architecture of functions and interactions working naturally in harmony. Specifically, observing animals, behavioural patterns become evident if you pay attention to them.
If you look at a flock flying over you, rapidly you’ll detect flight patterns. They do not fly randomly as an absent-minded could think. In their collective lives, there is a clear logic of movement that coordinates everyone in the group.
Back to birds and bees, the collective patterns that move the flock as a decentralized, self-organized system are called swarming intelligence. Swarming intelligence is perceived as a high level of synchronization amongst members whose the objective is to accomplish a task, collectively.
Swarming Drones: a date between bots
We said we have a pedagogical relation with nature. By using creativity and recent technology developments, the collective patterns we see in nature can be applied to drone tech. Multiple aerial machines can now interact and collaborate to achieve complex missions. For example, drone swarms could help in humanitarian missions by delivering food or medical drugs. Drones swarms could also complement firemen brigades in assessing wildfires and helping in containing them. They could help in disaster relief. Or even interact with humans in a play, providing smart surveillance or being part of a paintball game. Applications are endless. Machine to machine dialogue is reaching new levels.
But how?
Basically, by means of a complex hardware and software system running computational algorithms for communication, localisation, trajectory planning and collaboration. Current state-of-the-art drone tech requires multiple specialists working together to generate autonomous machine-to-machine interaction.
Nature has always inspired us. Drones can collaborate and interact harmoniously if we program them with a specific type of algorithm, the bio-inspired one. The metaheuristic we see by paying attention to the swarming behaviour in birds, ants and bees, for example, can be applied to make drones collaborate to solve complex problems.
Lessons from ants and birds: PSO and SDS algorithms
The interaction we observe in animals allows us to build mathematical models that solve big challenges. However, its application does not end there. In fact, one of the main uses of these algorithms is to explore new ways to develop drones collaboration tech.
We highlight two examples for bio-inspired algorithms: The Particle Swarm Optimisation (PSO); and the Stochastic Diffusion Search (SDS). Both are based on animal pattern behaviours that help analyze and solve problems. PSO was designed by observing flocking birds whereas SDS, on the other hand, has its origin in ants pattern behaviour analysis. Although both are based on natural being patterns, they serve different purposes.
Swarm Drones, a new social being?
The development of swarming intelligence algorithms applied to drone tech reflects a big step forward in the machine-to-machine interaction. It means drones have, in a particular perspective, the ability to socialize amongst them.
Of course it’s a little rushed to speak about “socialize” or even more of a “drones society”, but if we think about the possibilities that arise from combining, for example, computer vision, machine learning and obstacle-avoidance with swarming intelligence just to name a few, drones socialization becomes a potential reality.
Is this the birth of a new aerial social being?
