In order to solve this dilemma, Google has recently implanted in the online balloon artificial intelligence and machine learning system, AI technology, although the introduction of the physical level can not solve the high-altitude airflow changes, but can make high-altitude balloons to deal with this unstable change , To achieve the goal of long-term stay.
Compared with the previous static algorithm, there is AI technology to support the Project loom to the environment around the device to adjust the system to adjust the specific functions. From the publicity of the test case, this “online balloon + artificial intelligence” combination does have a strong ability to adapt, through data acquisition and analysis, intelligent regulation of height and bias, a balloon in Peru over a long time to stay up to 98 day.
Another balloon in the test analysis that the wind on the land is not enough to make it stay in a position, so simply temporarily floating over the Pacific Ocean. After making a pre-determined judgment about their behavior, it can still make rapid adjustments if there is a temporary change in weather.
After the introduction of AI, Google in the same geographical area can launch fewer online balloons, these balloons can not only stay in a place longer, but also reduce operating costs. To some extent, the introduction of AI technology also addresses the high cost of online balloon pain points.