The US military is set to create balloons that could float in the stratosphere in exactly the same spot indefinitely. The Defense Advanced Research Projects Agency (DARPA) - which has started testing the balloons - believes they could be used as cheap alternatives to satellites. Applications could include communications with remote areas or disaster zones, as well as surveillance of other nations. The solar-powered craft are able to anchor themselves in place thanks to sensors that can predict changes in the direction of the wind and an on-board motor compensates for the movement. Military aircraft fly at a maximum of 65,000 feet (20,000m) while these balloons operate at up to 90,000 feet (27,000m) and would therefore be virtually impossible to intercept.
High above the annoyances of weather and commercial air traffic, the stratosphere could be a great place from which to beam down Internet connectivity to places with poor communications infrastructure. Alphabet and Facebook are both working on drones to operate 18 kilometers or more above Earth, and this year Alphabet will start using balloons at that altitude to serve mobile subscribers in Indonesia. But even the stratosphere, which at the equator starts at around 20 kilometers but varies by latitude and season, is within reach of Earth's regulators. To work at large scale, Alphabet and Facebook's schemes will need significant changes to national and international rules. "This is all somewhat uncharted territory," says Yael Maguire, engineering director at Facebook's connectivity lab, which is working on a drone called Aquila that has the wingspan of an airliner (see "Meet Facebook's Stratospheric Internet Drone").
Google's parent Alphabet is set to beam internet to the remotest areas of the planet via high-altitude balloons. The firm has launched six balloons as part of its'Project Loon' that have managed to transfer data across a 620-mile (1,000km) area as part of a landmark test. A spokesperson from Loon, which is a subsidiary of Alphabet, said the stratospheric balloons rely on a single connection to the ground in Nevada. The test is Project Loon's latest as it heads towards its planned commercial launch of the service next year. Google's parent Alphabet is set to beam internet into the remotest areas of the planet as part of its'Project Loon' starting next year.
Project Loon is using balloons such as this to set up an aerial wireless network for telecommunications.Credit: Loon The goal of an autonomous machine is to achieve an objective by making decisions while negotiating a dynamic environment. Given complete knowledge of a system's current state, artificial intelligence and machine learning can excel at this, and even outperform humans at certain tasks -- for example, when playing arcade and turn-based board games1. But beyond the idealized world of games, real-world deployment of automated machines is hampered by environments that can be noisy and chaotic, and which are not adequately observed. The difficulty of devising long-term strategies from incomplete data can also hinder the operation of independent AI agents in real-world challenges. Writing in Nature, Bellemare et al.2 describe a way forward by demonstrating that stratospheric balloons, guided by AI, can pursue a long-term strategy for positioning themselves about a location on the Equator, even when precise knowledge of buffeting winds is not known.
The ability to analyze and forecast stratospheric weather conditions is fundamental to addressing climate change. However, our capacity to collect data in the stratosphere is limited by sparsely deployed weather balloons. We propose a framework to collect stratospheric data by releasing a contrail of tiny sensor devices as a weather balloon ascends. The key machine learning challenges are determining when and how to deploy a finite collection of sensors to produce a useful data set. We decide when to release sensors by modeling the deviation of a forecast from actual stratospheric conditions as a Gaussian process. We then implement a novel hardware system that is capable of optimally releasing sensors from a rising weather balloon. We show that this data engineering framework is effective through real weather balloon flights, as well as simulations.