Driven by computing power, not Newton or newtons.
New Zealand’s Canterbury Plain is hosting Google’s latest idea-that’s-so-goofy-it-might-work, appropriately named Project Loon. Thirty high altitude balloons carrying data relay equipment were released to drift over Christchurch, generally heading east towards the telecoms starved Chatham Islands. The concept Google is testing is to put enough balloons into the air to create a fleet of atmospheric satellites that can talk to each other and to the ground, and relay Internet service to hard to reach places.
With one major exception, the technology behind it is reasonably well established. In the late 1990s several projects emerged that incorporated various aspects of Project Loon. Teledesic, for example, proposed to launch more than 800 small satellites into low earth orbit and ring the globe with Internet access. SkyTower, a project I worked on for southern California drone pioneer AeroVironment, was an attempt to use a solar powered, unmanned airplane to maintain station over an area, at about the same altitude Google is testing, and serve as a low hanging satellite.
Teledesic, like the far less ambitious but actually launched Iridium and GlobalStar constellations, foundered because of the astronomical capital cost of building birds that would last for years and then getting them into orbit. The limitations of solar cell and battery technology did the same for AeroVironment. Google’s balloons solve both problems: they’re cheap to build and launch, can be easily recovered and refurbished, and the solar panels only have to power telecoms equipment and controls, not propellors.
The huge challenge for Project Loon is to figure out how to manage a free floating balloon fleet by modelling wind patterns and navigating simply by going up or down to find the right airstream. And do it in way that maintains a usable telecommunications network architecture. That is a horribly complex computing problem: just the sort of thing Google is good at.