Watch Drone Swarm Fly Through The False Forest Without Damage

Soria’s group tried a new way to oppose a advanced form compared to five drones and eight obstacles, confirming their search. Elsewhere, the teams completed their mission in 34.1 seconds — the predicted end with 21.5.

Next came the real show. Soria’s group gathered a little Crazyflie photos used by researchers. Each was small enough to fit in his hands and weighed less than a ball, but he carried an accelerometer, a gyroscope, a small device, radios, and smaller ones. walk-walk balls, were between a few inches and between four leaves. Reading from a sensor with a camera in the living room, which monitors the balls, descends on a computer that drives each type of drone as a ground ball. (Small drones cannot carry the necessary equipment to be able to do the counting.)

Soria placed the drones down “in the beginning” near the original barriers like a tree. When launching the experiment, five drones woke up and quickly moved to a 3D platform at the point of departure. Then the copyists started walking. They circled the air, between the soft green barriers, passed through, on the ground, and around each other, and headed for the end where they slowly reached. No collisions. The smoothness of the possible increase is possible due to the mathematical calculations that are repeated in real time.

“The results of the NMPC [nonlinear model predictive control] these examples are very useful, ”wrote Gábor Vásárhelyi, a robotic researcher at the University of Eötvös Loránd in Budapest, Hungary, by sending an e-mail to WIRED. (The Vásárhelyi group developed the Soria type they used, but did not participate in the project.)

However, says Vásárhelyi, this research does not address the fundamental barrier to the development of a predictable future: reading requires a central computer. Adjusting long-distance adjustments can leave the whole team vulnerable to connections or errors. The only way to drive people away is not to find the best way to get through, but “they can operate very small weapons (like mosquitoes, small bugs or small drones) and grow well, the size of them,” he writes. The design — as well as the environment — drone teams may not have more computers.

“This is a question of opinion or quantity,” Vásárhelyi continues. “However, nature has it all.”

“That’s when I said ‘Yes, I can,'” said Dan Bliss, an engineer at Arizona State University. Happiness, who does not participate in Soria’s group, is leading the Darpa project to make mobile phones more efficient with the use of drones and consumer technology. Even small drones have to become overly sensitive over time. “I take a computer problem of several watts and try to put it on a processor that uses 1 watt,” he says. Excitement also adds that creating an independent show is not only a problem, but also a cognitive problem. The tools on the world map, such as computer screens, require a lot of energy.

More recently, Soria’s team has been working to distribute intelligence among drones to meet larger groups, as well as to overcome major obstacles. Eagles that love to predict, like drones sending a burrito, years later. But that is not the case no. Robots can see them in the future, and especially to their peers.

Many Great Stories