Produced by: MOS Video Department

On May 4, 2022, CalTech researchers published in the journal Science Neural-Fly technology, that a new neural-technology can be used in drone deliveries or flying cars. At present, the premise of UAV flight is ideal weather and an unobstructed environment.

However, this is difficult to ensure such environments in reality. Using deep learning method of the neural network, the CalTech research team designed the Neural-Fly self-learning ability to make it.

It can automatically adjust the program steps. The researchers also updated this technology’s key algorithms and parameters so that the drone can automatically adjust the flight according to environmental changes and wind conditions. The project was first flight-tested at Caltech’s CAST Laboratory in a wind tunnel. The flight was tested in the wild with strong winds, and the results were good. In addition, the CalTech team demonstrated drones for collecting and disseminating information and built a knowledge base for autonomous vehicles.

The new technology invented by the California Institute of Technology is expected to use aerial flight to solve the current global traffic congestion problem.

Posted by: Xueli