Alphabet’s internet air balloons, which beam down internet access to rural and remote areas of the world have relied on algorithms designed by humans to navigate ever since they first started to provide internet access over Kenya in July 2020.
Now, it looks as though the stratospheric helium balloons have a new navigation system in the form of an artificial intelligence system developed by Google AI, called Reinforcement Learning (RL). The RL navigation system is now in charge of managing Loon’s fleet of balloons over Kenya, which Loon says is the world’s first deployment of reinforcement learning in a production aerospace system.
The benefits of using the new RL navigation system is the fact that it can figure out the optimal route for ballons to travel much faster than algorithms designed by humans.
“While the promise of RL (reinforcement learning) for Loon was always large, when we first began exploring this technology it was not always clear that deep RL was practical or viable for high altitude platforms drifting through the stratosphere,” said Sal Candido, Loon’s chief technology officer.
“It turns out that RL is practical for a fleet of stratospheric balloons. These days, Loon’s navigation system’s most complex task is solved by an algorithm that is learned by a computer experimenting with balloon navigation in simulation,” Candido added.
In order to test the effectiveness of RL vs. StationSeeker, the older generation navigation system using an algorithm, the team at Loon and Google AI measured the proximity over time of the two respective balloons to one of Loon’s Romer boxes, which is a device we use to measure LTE service levels in the field. the results showed that the RL navigation system outperformed the conventional system, mitigating the drift of its balloon to consistently keep it closer to the Romer box throughout the entire test.
— New Scientist (@newscientist) December 2, 2020