Swarms of drones flying with “consciousness”

https://cdn.steemitimages.com/DQmcwXHqCGqveiREWuYUt7QdhqEBi3WjP7RBTUp1DMqdvcB/imagen.png

Inspired by nature.

Advances in flying drones are moving towards a point where they can act as if they were a single organism, avoiding obstacles at high speed, without sophisticated sensors, without communication between them and without any human control.

A team of researchers in Shanghai was inspired by the vision of fruit flies and with that they created a system that allows robotic swarms to navigate chaotic environments with almost instinctive precision using only physics, simulation and a camera that would barely distinguish someone's face up close - natural intelligence translated into algorithms.

https://www.youtube.com/watch?v=LKg9hJqc2cc

And most impressively, all of this works on an insert that costs only $21.

The innovation starts from a simple but disruptive principle, eliminating the traditional complexity of navigation systems for drones, instead of dividing the tasks into mapping, detection, planning and control, the Chinese researchers replaced that entire architecture with a single compact neural network trained from start to finish.

This network learns to fly using differentiable physics, a technique that allows simulating movements and feedback to the system to adjust commands. This drastically reduces the need for real data, accelerates learning and eliminates dependence on expensive sensors. The system uses a camera of just 12x 16 megapixels, a resolution so low that it is reminiscent of the segmented vision of insects. Even so, it manages to interpret the environment and guide the drones between obstacles as if it were instinctively programmed for that.

This model works in a self-organized way, that is, the drones do not need to communicate with each other, wait for orders from a central system, each one interprets the environment autonomously, but they still move in a coordinated way, as if they shared a silent collective consciousness, this guarantees a quick and efficient response even in unpredictable environments such as disaster zones, collapsed buildings or dense forests.

By eliminating bulky sensors or high-definition cameras, the system significantly reduces weight and energy consumption, increasing autonomy and reducing the risk of failure.

References 1

Follow my publications with the latest in artificial intelligence, robotics and technology.
If you like to read about science, health and how to improve your life with science, I invite you to go to the previous publications.
You want to win, play HARRY-RAID