Project Loon uses artificial intelligence to fix the location, catching the wind to fly high

Project X's Development Loon team (Alphabet, Google's parent company) has revealed that they are using artificial intelligence (specifically machine learning) to help position the wi-fi -Fi in the long run.

X's test shows that in the Peruvian airspace, the balloon can remain in place for 98 days even with strong winds and steady turns. Google's previous solution was to use static algorithms to hold positions, but this approach has many limitations and can not be changed on a case-by-case basis. while airships rely on the flow of air to fly.

Flight path of the balloon.

Artificial intelligence helps keep the balloon in place for longer periods of time.

In addition to using machine learning , Wired says the balloon constantly collects weather data and learns from them. In one case, the balloon can know when the wind is not strong enough to keep the system floating in the sky. At this point, the balloon will temporarily fly toward the Pacific Ocean to catch the wind. The balloon can even improve its own behavior after predicting what is going to happen. Within 14 days, the balloon adjusted itself 20,000 times during its journey from Puerto Rico to Puerto Rico.

The new artificial intelligence not only keeps the balloon in place for longer, but it also helps reduce costs and increase project impact. The company will not need to use too many balloons to broadcast Wi-Fi for the same area, or simply keep the number of airships but expand to more places.