Elon Musk's spaceship company SpaceX is launching tens of thousands of small communications satellites. You might be wondering what it means for shipping.
This article is an attempt to bring together some of the commentary and facts to put together a picture. But please note that very little information has been released about the company’s market intentions, so there is a lot of conjecture in this article.
SpaceX was founded in 2002 by Elon Musk, with a goal to make it possible to colonise Mars. The satellite constellation is something it thought it could launch additionally, at a cost of just $10bn (2018 estimate).
By owning its own rockets, SpaceX can launch a satellite constellation much more cheaply than other satellite companies, who have to make a contract with a separate rocket launching company.
SpaceX’s plan is to launch 60 satellites every 2 weeks from early 2020, starting on Jan 7 and Jan 24, 2020. It follows 2 x 60 satellite test launches in May and November 2019, and 2 test satellites launched in February 2018. 60 satellites can be launched at once from the company’s “Falcon 9 Block 5” rocket.
At this rate of launching, it will have enough satellites in the sky to start offering a service in mid-2020, and for global coverage by the end of 2020 (after 24 launches), and 12,000 satellites to be deployed by mid 2020s, providing more and more capacity.
The company said it was developing a new rocket called “Starship” which could launch 400 satellites at a time.
Other than that, there isn't a lot of public information about Starlink. We know about the launches, we know about the filings made with the various regulatory bodies, and there is the odd tweet from Elon Musk.
But then, there's an enormous amount of commentary on the internet about what it might be, what it might offer, and if it might work financially. Some of this commentary comes from highly knowledgeable and well-placed sources, including the former president of the Mobile Satellite Users Association; a NASA engineer who formerly worked on Hyperloop One; and a University College London Professor of Networked Systems. Some links to expert blogs and videos are at the bottom of this article.
For shipping
A first point to consider - from shipping's point of view - is that data to and from satellites goes in straight lines.
This is fine for geostationary satellites, such as those operated by Inmarsat, which are 35,786 km from mean sea level. The data can go from the ship to the satellite and back to the ground station, and reach a ship in the middle of the Pacific.
But for low earth orbiting satellites like SpaceX (or Iridium), the satellites are too close to the earth’s surface for the communications to go from ground station to satellite to ship in one hop, if the vessel is far out at sea.
Iridium solved this problem with a complex system for sending data from satellite to satellite. But Iridium does not operate at gigabyte bandwidths. Its maximum speed is 704kbps.
SpaceX has announced plans to build a system for communications from one satellite to another by laser. Laser based communication is already used on spacecraft.
In an October 2019 interview, CEO Gwynne Shotwell said, “By late next year, we'll be flying satellite with lasers that allow them to talk to each other in space and share data, which ensures customers will never lose service”.
But this laser-based system does not exist on the satellites being launched during 2020. Whether this will set-back the systems’ applicability for shipping for a year, or longer, will depend on how successful SpaceX is at developing it.
It does sound quite tricky technically to have a laser on one moving satellite pointed with complete accuracy to another moving satellite, but it may be feasible.
Without intra-satellite communications, the service can only work within a certain distance to a base station. We can calculate this distance using trigonometry and the angle of elevation.
In a March 2017 FCC filing, SpaceX said that its constellation was designed to provide service at “minimum operational elevation angles of 35 degrees”, in order to minimize the effect on terrestrial systems.
If the satellite is 340 miles above the earth, and the minimum elevation angle from the ship to the satellite and back to another base station is 35 degrees, we can calculate there must be another ground station within 970 miles for it to work.
An alternative is that SpaceX builds its own base stations in the ocean. If a customer satellite terminal can also act as a base station, taking data from one satellite and transmitting it to another, these base stations could be the satellite terminals fitted on ships of other SpaceX customers.
Or SpaceX could position its own ships across the ocean as base stations, which sounds ambitious until you remember this is a company which built its own rockets.
The same issues would apply if providing services to aviation over the sea.
Another issue to consider is how maritime satellite communications services today involve much more than airtime. Supplementary services offered by maritime satcom providers include separate accounts for crew members, e-mail, cyber security services, technical support and software.
Also Inmarsat and Iridium communications services come backed with a safety guarantee (in order to be accepted as GMDSS equipment), and reliability of data transmission can be more important than the price.
So it may require more than much lower cost airtime and faster speeds to encourage shipping companies to switch away from traditional providers.
Main markets
But also bear in mind that SpaceX has never shown any interest in the maritime market.
It has not said anything much in public, but comments from the CEO last October hinted that the market they have in mind initially is US customers in remote places, who often still pay $80 a month for fairly poor quality broadband delivered through cables.
Another anticipated market is “backhaul” internet traffic, replacing fibre optic cables.
Providing satellite communications services to this market does sound plausible, if SpaceX can install a base station across the terrestrial US every 970 miles. The base stations do not need connection with fibre optic cables, they just need to have power to receive signal from one satellite and send it up to another.
Ground stations
Another question is the cost of the user terminals, the device which customers install to send and receives data to and from the satellites. Elon Musk tweeted in Jan 2020 that it looks like a "UFO on a stick" – which users would just need to plug in and point at the sky, with no training required.
Experts have said it is probably a phased array antenna, a device which monitors the location of satellites and sends data communications directly to them, as they move.
SpaceX has said it believes the device can be made for $200. News agency CNN interviewed Bill Milroy, chief technology officer of US antenna maker ThinKom, which makes computer controlled phase array antennas. He said that if he was asked to make user terminals as cheaply as possible but at large volumes, it might be possible to get the cost down to $1000 each.
Mark Handley from University College London, has speculated that the user terminals could be ground stations themselves, handling data from other customers, not just their own. The data would be sent from a satellite to a user terminal, and then up to another satellite.
The initial terminals are being made by SpaceX itself.
Direct-to-customer
The communications services are likely to be sold directly to customers, similarly to how Elon Musk’s Tesla cars are sold to customers directly. Although in some countries, regulations will require that the service is offered through local telecom companies.
The company will be hiring a whole workforce for sales, technical support and product engineering, looking in particular at user terminals.
It may initially target US customers who are paying high fees for poor quality broadband. CEO Gwynne Shotwell mentioned "millions of people in the U.S. pay $80 per month to get “crappy service.”
Any customer with access to reasonably fast broadband is probably not going to be interested in this - whether cabled or via fast cellular. So it will probably not be of interest to customers who have good broadband or cellular coverage.
Other issues
Other issues commentators have raised is the challenges with the “very low earth orbit” altitude, because there is atmospheric drag, which would normally result in a short lifetime for the satellite. The information about SpaceX’s technology for this has not been made public.
To handle the “space debris,” the company announced in late 2017 that satellites nearing the end of useful lives (5-7 years) would move into a ‘disposal orbit’, and re-enter the earth’s atmosphere.
SpaceX has a contract for $28m with the US Air Force Research Laboratory, to test way that the services can be used for military.
The company is also in a race with other companies planning similar constellations, including OneWeb.