Facebook Ditches Aquila

aquilaFacebook recently announced that they will be stopping their Aquila drone initiative, instead relying on other companies to build high altitude aircraft. In a company blog post, Facebook said that they no longer plan to build their own equipment since the broader industry is now interested in the concept.

Facebook continues to support connectivity programs for the ~four billion people currently without internet access, including fiber programs, terragraph, and policy initiatives such as a proposal for 2019 World Radio Conference to get more spectrum for High Altitude Platform Station (HAPS) systems. Facebook also is quietly investing in a next generation satellite program.

Celebrating Fiber

fiber.jpgAs broadband expands throughout the globe, most communications are still carried by fiber optic cable. Can we take a minute to celebrate the marvel of this technology?

Scientists have known for 150 years that glass can be used to guide light. In the 1980s, manufacturers improved techniques to make highly transparent threads of glass the width of a human hair and over a hundred kilometers long. Simultaneously laser technology was getting cheaper and smaller, and digital data processing getting faster.

So why bother with fiber instead of traditional copper (the first undersea copper cable was laid across the Atlantic in 1858)?

For starters, fiber optic cables have an unbelievable capacity for carrying information. A single fiber, for example, can carry 3,000,000 simultaneous phone conversations. Since a cable can comprise over 1000 fibers, this means a single cable could support three billion conversations — or half the planet speaking with the other half, simultaneously.

Light travels efficiently with very low attenuation. Signals can maintain sufficient strength for over 100 kilometers before needing a boost.

Cables carrying information with pulses of light aren’t subject to electromagnetic interference the way typical copper cables are. The signals avoid corruption (and eavesdropping is much more difficult).

And one more important characteristic of fiber optic: the main ingredient in a cable is silica (aka sand). While copper cables around the world are highly prone to theft (copper can cost a few dollars a pound, and large cables will weigh tons), if thieves want silica, it’s a lot easier to pilfer the beach!

Update on CubeSats

cubesatCubeSats are miniaturized satellites which comply with agreed to standards, including component cube dimensions of 10 cm on a side and less than 1.3 kg of weight per unit. Imagine a container with a liter of water — that is about the size and weight of a CubeSat.

Because they are so small and primarily use commercial off-the-shelf components (mostly designed for cell phones), CubeSats are fast and cheap to design and deploy. Historically they have been launched as secondary payloads with larger launches. Over 800 CubeSats have been deployed to date, and at least 1200 more are planned for orbit. A new industry of launch services targeting CubeSats (and other small satellites) is taking shape.

The simplicity and low costs of CubeSats means many groups can now become involved in space science. Universities, high schools, and individuals have all designed and launched CubeSats. Some have even been funded by KickStarter campaigns.

Developing countries are also involved. For example, Kenya recently designed the CubeSat 1KUNS-PF which was carried to the International Space Station by a SpaceX resupply mission, and from there launched into orbit. Over 18 months it will assist with mapping of Kenya, monitoring the coastline, and identification of illegal logging. To date, an impressive 80 countries have launched CubeSats.

So to summarize, 800 CubeSats have been launched by 80 countries, with 1200 more already scheduled to go!

The Concern of Space Junk

space junk.jpgThere currently are less than 2,000 operational satellites in orbit. In the next few years, SpaceX and other launch services are going to be deploying tens of thousands of new satellites. What does this portend for the problem of space junk?

The short answer is that it is hard to know for sure.

The good news is that space is big, including even near-earth orbit. There is a lot of room for a lot of satellites. If you think, for example, about the number of boats the oceans can accommodate, and then realize that space is much, much larger, it gives an idea that there is a lot of room to work with. The odds of a collision are low.

All new satellites need to have launch approvals and also decommissioning plans (typically involving falling back into the atmosphere and burning up). SpaceX and OneWeb, for example, have committed to one year deorbiting plans for satellites at end of life.

We’re also pretty good at tracking larger pieces of space junk and identifying potential problems. The International Space Station is periodically moved in order to minimize chances of collision. Around 20,000 man-made objects are currently tracked in space (although they need to be big enough to track — estimates assume many millions of smaller items are also in orbit).

The concerning problem is that one collision can lead to the creation of lots more space junk, which in turn could collide into other satellites. Computer models show that a chain reaction of this sort is possible (the “Kessler Syndrome”). This also isn’t hypothetical: at least five satellite collisions have resulted in increased space debris. Both the International Space Station and the Mir Space Station have sustained damage from collisions with space debris. Space crowding is particular acute at the poles: many satellites maintain polar orbits (in order to have complete coverage of the earth), which means the orbits all cross at the poles.

Some scientists argue we are already in the early stages of the Kessler Syndrome.

So on the issue of space debris, most scientists cautiously believe we are OK if we are prudent. But there are definitely unknowns, and definitely risks.

Facebook Gets Into the Satellite Business

fFacebook has reportedly registered a new subsidiary to build low earth orbit (LEO) satellites, competing with SpaceX, OneWeb, and others. The subsidiary, called PointView Tech, plans to launch a demonstration satellite in 2019 to investigate using the E-band spectrum for communications. E-band promises much higher data connection speeds than those planned by rivals, but needs to overcome challenges, including absorption by rain or other particles. E-band is also used by the Facebook drone project called Aquila.

For the Facebook satellite constellation to work, there would need to be thousands of satellites, similar to SpaceX and OneWeb.

The PointView Tech initiative puts Facebook in direct competition with SpaceX. There doesn’t appear to be much love lost between Mark Zuckerberg and Elon Musk. They have engaged in a public feud around AI. Musk recently deleted all Tesla accounts from Facebook. The relationship also wasn’t helped when Facebook’s last satellite project, AMOS-6, blew up on launch of a SpaceX rocket in August 2016.

Low Cost Satellite Networks

cubeWhile SpaceX, OneWeb, O3B and other multi-billion dollar satellite constellations garner most of the press, other lower cost initiatives demonstrate a different and potentially consequential approach.

Sky and Space Global, for example, plans to launch 200 nano-satellites (under 10 kg each) into low earth orbit in order to provide telecommunications services in Africa, Latin America, and elsewhere. The satellites, which adhere to CubeSat standards, will be deployed in near-equatorial planes, reaching 15 degrees north and south of the equator.

Satellites will be launched aboard LauncherOne, the air-launched rocket from Virgin Orbit. Satellites will communicate with ground antennas which provide wifi hotspots, or potentially with a new generation of $20 Android phone capable of direct communications with the satellites.

Sky and Space Global aims to build and launch the entire constellation of 200 satellites for $200 million, a fraction of the cost of even one geosynchronous communications satellite.

LandCruiser Emergency Network Project

cruise70% of Australia lacks cell coverage. Even remote areas, however, do boast lots of Toyota LandCruisers crisscrossing the terrain.

Flinders University, along with Toyota and Saatchi & Saatchi Australia have proposed outfitting LandCruisers with communications hubs capable of “store and forward” messaging. Each “mobile hotspot” would include wifi, UHF and mesh networking capabilities with a range of 25 km. Messages would be passed from vehicle to vehicle until reaching an internet-connected base station.

The LandCruiser Emergency Network wouldn’t provide true broadband, but would offer messaging services, especially useful during emergencies.

Low Earth Orbit and Steerable Antennas

satThe advantage of geosynchronous orbit is that satellites appears stationary. Satellite dishes or antennas tracking the satellite don’t need to move. Any orbits other than geosynchronous require antennas to move to track the satellite. Historically, this added a lot to the complexity and cost of the antenna (although the Soviets employed the “Molniya Orbit” for decades which required dishes to nod up and down from the horizon).

As companies contemplate placing thousands of satellites into low earth orbit, and all of the advantages that confers (less latency, smaller satellites, lower cost), a major challenge appears: How do you design an antenna to track satellites, including frequent handoffs from one satellite to another? And if the antenna is moving in a plane or car, how does that factor in?

Fortunately, there is great progress in a new generation of “steerable antennas”, also described as a “phased array antennas”. Researchers have essentially built the “steering” elements, until now managed through motors, onto a chip. Flat panel antennas are being designed which can track satellites, including through the frequent passing from one to another.

The technology is well-demonstrated, and a number of agreements are being signed between antenna technology firms and satellite companies, such as recent agreements between ALCAN and SES or between Phasor and LeoSat.

Technology firms are still wrestling with costs for steerable flat panel antennas, although with millions likely to be purchased for broadband access, companies are optimistic that prices will fall to a few hundred dollars.

Bridge International Academies

bridgeHundreds of millions of children have no school to attend. Hundreds of millions more attend schools with poor facilities, minimal supplies, and frequently absent teachers.

Because of this dire situation, a number of countries are experimenting with international private schools that focus on new technologies, broadband linkages, standardized curriculum, rigorous evaluation, and low cost.

The best known of these is probably Bridge International Academies, whose high profile is partly due to an august list of investors, including the Gates Foundation, Chan Zuckerberg Initiative, Omidyar Network, and World Bank. Bridge currently operates in five countries: Kenya, Uganda, Nigeria, Liberia, and India. Over 100,000 students attend one of more than 500 Bridge schools. Bridge aims to educate 10,000,000 pupils by 2025.

Bridge provides teachers with a tablet that includes all lesson plans in highly scripted formats. Bridge rigorously collects data about teacher and student progress. Administrative cost are kept low due to centralization of many tasks; each school requires just one administrator with a smartphone app. Costs for students depend on region and economic status. In Uganda, for example, parents pay about $66 per year, which is much cheaper than other private schools and roughly on par with “free” public schools that require a number of purchases.

Bridge points to studies which demonstrate that its students out-perform public school children.

Simultaneously, private school networks — and Bridge International Academies in particular — are lightening rods for an exceptionally high level of controversy. Bridge has had periodic conflict with ministries of education, teachers unions, and other organizations with strong opinions about education.

Location Services in Developing Countries

locationWhere am I?

Answering this most basic of questions can represent a major challenge in developing countries. In regions with no maps, no addresses, sometimes no names, it is difficult to know location. And without location, it is impossible to meaningfully engage with the rest of the planet.

New technologies offer powerful solutions regarding location services.

First, and most fundamental, is global mapping. Google Maps, ESRI, and other services offer detailed traditional and satellite view maps. When conventional maps don’t exist in a location, researchers can now easily add them. For example, vaccine researchers at the Gates Foundation  analyzed satellite images for regions not yet immunized — often because of inaccurate maps — in order to build accurate vaccination plans.

Second, new technologies can help define property rights. Over a billion households still live without property rights to their homes that are secure, registered, documented and tradable. These “hidden” rights are economically significant — likely exceeding $10 trillion in value. New registries are helping. The World Bank and others have invested in open cadastre systemsDrone technology can play a role. Even distributed blockchain technologies may become increasingly useful.

Finally, how does one describe their location if no addresses exist? By providing GPS coordinates of two nine digit numbers? A British firm called what3words has a clever solution. They divide the planet into a grid of 3 meters x 3 meters and have assigned each square a unique three word identifier (I’m currently writing, for example, from this beautiful corner of the planet: searching.colonialist.suggested).