Since Google introduced the Alpha Dog, the artificial intelligence program has been popular, and the Alpha Dog is the first artificial intelligence program to defeat the World Championship of Go. According to reports, Google has pushed "hard" technology, this time is the AI ​​drone.
Google, which has always had strong software genes and lack of hardware strength, has jointly funded NASA's Jet Lab (JPL) and leveraged JPL's hardware capabilities to jointly develop fully autonomous small drones. Can't help but guess, Google is going to make Alpha "chicken" after Alpha "dog"?
JPL combined with Google's 3D environment detection technology Tango, recently built an AI-controlled drone, and recently held a drone indoor test flight with human and AI operations. Although in the game, the AI ​​drone slightly fell behind the 11.1s of the human-controlled drone with a score of 13.9s per lap. In contrast to human manipulation, artificial intelligence algorithms rely more on sensations during flight. The AI ​​drone will be more stable during the flight, while the human pilot tends to accelerate aggressively and the flight route is more aggressive.
In particular, manually operated drones typically rely on GPS for navigation, but this limits their flight in indoor spaces. Therefore, in warehouses or intensive cities, AI-controlled drones have an advantage because they can break the GPS limit. At the same time, the AI-driven drone is also a completely autonomous drone, which means that the drone has no external positioning and external computer control, and simply relies on its own judgment to select the optimal route to fly.
In addition, the maximum performance limitations of fast indoor flight are mainly due to the shutter response speed of the onboard camera, which is used to track the movement of the drone. Flying too fast, too close to the ground, or rolling and pitching too fast can result in blurred images and even loss of drones.
JPL solves this problem in two ways: First, using two wide-angle cameras, one pointing to the front and the other pointing down, so that the drone can see the horizon in a field of view greater than 250 degrees. Second, adjust the flight path of the drone to limit the rate of rotation and the ratio of speed to altitude.
At present, the main limitation of the development of AI drones is camera technology and artificial intelligence algorithms. Camera technology can give the drone a pair of "eyes" to keep up with its flight movement; the improvement of artificial intelligence algorithm is like giving it a stronger "brain", let it learn in the shortest time. Analyze the optimal flight path. With JPL's hardware advantages, combined with Google's own software experience, I believe this AI drone will surpass the human-operated drone.
In summary, with the development of AI technology, AI drones will play its role in a wider world. In the future, in addition to the indoor environment, in the search and rescue of deep mountains and wild forests, in the flood support, there may be AI drones.
Copper Tube Terminals Without Checking Hole
Our company specializes in the production and sales of all kinds of terminals, copper terminals, nose wire ears, cold pressed terminals, copper joints, but also according to customer requirements for customization and production, our raw materials are produced and sold by ourselves, we have their own raw materials processing plant, high purity T2 copper, quality and quantity, come to me to order it!
Copper Tube Terminals Without Checking Hole,Cable Lugs Insulating Crimp Terminal,Cable Connector Tinned Copper Ring Terminal,Tubular Cable Lugs Crimp Terminal
Taixing Longyi Terminals Co.,Ltd. , https://www.longyicopperterminals.com