1. Elon Musk’s vision…
Elon Musk is one of the most divisive, disruptive, creative, and successful entrepreneurs of the 21st century – and will gift the future with either a step change in technology, or with endless cautionary tales and missed opportunities.
One potential step change is if everyone on Earth had access to broadband internet, and the rising tides of education, business and public service such access would bring. Musk started Starlink in 2015 with that step change as a goal.
What is Starlink, and how did it come about?
Musk tells the story that when he was in college he thought about what technologies he was interested in that might most affect the future of humanity. His list included renewable energy, the internet, and space colonization (or making humans a multi-planetary species).
Musk’s first forays into the internet became what is now Paypal with a current market value (as of August, 2020) of $240B. Elon’s efforts in renewable energy has become Tesla with a market value of about $440B. And Elon’s interest in making humanity a multi-planetary species has become SpaceX, valued at around $42B.
One of SpaceX’s goals is to colonize Mars. In order to colonize Mars, a subsidiary goal was to make rocket transport as cheap as possible. For rockets to become much more affordable, they must be reusable. Reusability has driven SpaceX’s technological efforts to date, specifically to launch and land their rockets so the majority of the structure can be reused again and again like an airplane. Their technical successes in meeting this goal has enabled SpaceX to launch rockets at significantly lower costs than the competition, and now SpaceX dominates the global launch business.
Having a low-cost launch provider in SpaceX has opened up the opportunity to re-address one of Musk’s three high-impact technologies which was the internet. That has become the new goal: to provide low-cost high-speed broadband internet to all corners of the globe, especially to people and places that are deprived of services now.
In addition to the humanitarian motivations for providing global broadband internet, is the more prosaic needs of the Mars mission. Rockets are a highly capital-intensive business. And Musk has stated that satellite internet “is intended to generate a significant amount of revenue and help fund a city on Mars”.
That has led to a new company called Starlink, a division of SpaceX.
2. A brief history of Starlink…
In order to provide global broadband internet, Starlink is currently launching a constellation of very low-earth orbit (VLEO) satellites. The constellation may consist of up to 42,000 small satellites, according to recent FCC filings, in orbits that range from 328 to 580 to 1,200 kilometers in altitude. That’s where things stand today – but it was not always so – and may change in the future.
In a private event at the Seattle Center in January, 2015, Musk announced the launch of a yet-unnamed satellite internet company. At the time, Musk estimated that it would cost $10 billion to launch as many as 4,000 satellites orbiting at an altitude of about 1,200 km.
Starlink began product development of the satellites in Redmond, WA in 2015, and initially aimed to launch two prototypes by 2016. The prototypes were ready in 2016, but launches were delayed due to a dramatic September, 2015 launchpad explosion. Furthermore, Starlink was struggling with the economics of the user terminals, aiming for an easy-to-install system a user could set up for $200.
SpaceX filed plans with the FCC In March, 2017, to expand the constellation with over 7,500 additional satellites orbiting at the VLEO altitude of around 340 km. Later that year, SpaceX filed documents to trademark the name “Starlink”, inspired by the book “The Fault in Our Stars” by John Green.
Finally, the two prototypes named Tintin A and B were launched on February 22, 2018. They were secondary payloads on a SpaceX Falcon 9 rocket whose primary payload was the Spanish Paz Earth-imaging radar satellite.
In March, 2018, SpaceX gained conditional FCC approval for the first 4,425 satellites at the higher non-geostationary orbit (NGSO, above the international space station, ISS).
Musk traveled to Redmond in June, 2018 to fire several of the Starlink senior management team including Rajeev Badyal, formerly of Microsoft and Hewlett Packard, and Mark Krebs, formerly of Google. Musk was reportedly unhappy about the pace of development. Please note the expectations given the performance: a little over two years from January, 2016 to March, 2018 for the launch of two groundbreaking new prototypes, and Musk was unhappy. Of course, he may have been even more unhappy about the projected future development schedule. There is no question Musk is a demanding boss.
In November, 2018, the FCC approved the larger constellation of 7,518 additional satellites at the lower VLEO (below the ISS).
In April, 2019, the FCC approved SpaceX’s request to modify the constellations, and the end result was the nod for a total of about 12,000 satellites distributed in three different orbital shells, the original high NGSO, the low VLEO, and an intermediate altitude of 550 km. This was also the month that SpaceX transitioned from predominantly satellite research and development to manufacturing.
On May 2019, SpaceX launched the first 60 of Starlink’s nascent constellation in what was going to be “only” 12,000 satellites. Since then, SpaceX has been regularly lofting 60 satellites at a time on their Falcon 9 rockets. As of August, 2020, Starlink’s constellation has grown to 653 satellites (also marking the 6th re-use of a Falcon 9 first stage rocket in that launch). Starlink is currently manufacturing 120 satellites per month.
In December, 2019, the FCC approved SpaceX’s request to further modify the number of satellites (tripling the number of orbital planes) in each orbital shell, to bring satellite coverage to the southern US in time for the 2020 hurricane season. Think of the orbital shell as the surface of a ball. The orbital planes are all the maximum diameter circles you can draw on the surface of the ball.
The goal for 2020 is to launch 60 satellites every few weeks, and to become operational in regions lacking reliable and affordable broadband internet in some rural areas of the US.
3. Competition…
SpaceX/Starlink is one of many companies trying to provide satellite internet. Other companies include OneWeb, Amazon, Samsung, Boeing, Hongwan, and many others.
OneWeb was founded by American engineer and entrepreneur Greg Wyler and was formerly known as WorldVu Satellites.
The first batch of OneWeb satellites was launched in February, 2019 aboard a Soyuz rocket launched by European company Arianespace. The plan was for Arianespace to launch 21 rockets carrying a total of 900 OneWeb satellites.
OneWeb filed for Chapter 11 bankruptcy in March 2020, and laid off 85% of its 531 employees after its biggest investor, Softbank, declined to provide further funding. OneWeb’s biggest creditor was Arianespace, who was owed $238 million. OneWeb had signed a $1.1 billion contract for those 21 launches.
Although we don’t know whether launch service costs drove OneWeb into bankruptcy, it was probably one of the biggest expenses which highlights SpaceX’s fundamental competitive advantage. Let’s dig into that a little.
We know that SpaceX is the low-cost launch provider. The average Falcon 9 launch cost is about $60 million with a capacity of 8,300 kg to a geosynchronous transfer orbit (GTO). The Arianespace Soyuz has a capacity of 3,250 kg to GTO.
Some simple math shows that 21 launches gives a total project capacity of 68,250 kg aboard the Soyuz rockets, for a total price of $1.1 billion charged by Arianespace to its customer OneWeb.
SpaceX could have launched all of OneWeb’s constellation in eight or nine Falcon 9 launches for at most $558 million, or about half the price paid to Arianespace. Even better, the Falcon Heavy has a capacity of 26,700 kg to GTO and a launch cost of $90 million each. OneWeb could have lofted two or three Falcon Heavies, for a total cost of $270 million, or about a quarter of Arianespace’s cost of $1.1 billion.
On top of that SpaceX charges a profit margin to their customers, which means the same launch services costs Starlink much less.
OneWeb was between the proverbial rock and a hard place: pay their competitor SpaceX for affordable launch services and essentially bankroll Starlink’s efforts and shoot themselves in the gut, or pay exorbitant prices and go bankrupt and shoot themselves in the head. Gut or head?
Most of SpaceX’s satellite internet competitors are facing the same tough competitive dynamic that crumpled OneWeb. So, we won’t spend any more time with them.
But there are a couple other competitors who can potentially provide in-house launch services for their own satellite divisions.
One is Boeing, who has a joint venture with Lockheed Martin called United Launch Alliance (ULA). One of their workhorse rockets is the Atlas V, which costs $110 million in 2016 to launch a maximum of 8,900 kg to GTO. Almost double the cost of SpaceX’s Falcon 9 to launch a roughly similar payload. ULA has a larger Delta IV rocket which costs about $164 million to loft up to 14,220 kg to GTO. The SpaceX Falcon Heavy can throw almost double the weight that a Delta IV can, for a fraction of the cost.
The last competitor worth mentioning is Amazon. Amazon’s founder and CEO, Jeff Bezos, has a private venture called Blue Origins, which is similarly developing reusable rockets. As of now, Blue Origins is unable to reach orbital altitudes, and does not provide total cost or payload data. Blue Origins advertises flights up to 100 km. VLEO orbits, the lowest at which Starlink is deploying satellites, is 328 km.
4. The shiny new toys…
The Starlink satellite is described by SpaceX as follows in one of their presskits:
“Each Starlink satellite weights approximately 260kg and features a compact, flat-panel design that minimizes volume, allowing for a dense launch stack to take full advantage of Falcon 9’s launch capabilities. With four powerful phased array antennas on each satellite, an enormous amount of throughput can be placed and redirected in a short time, for an order of magnitude lower cost.
The system is on the leading edge of on-orbit debris mitigation, meeting or exceeding all regulatory and industry standards. At end of their life cycle, the satellites will utilize their on-board propulsion system to deorbit over the course of a few months. In the unlikely event their propulsion system becomes inoperable, the satellites will burn up in Earth’s atmosphere within 1-5 years, significantly less than the hundreds or thousands of years required at higher altitudes. Components of each satellite are fully demisable.”
The phased array antennas are a much simpler, quicker and cheaper way to steer a beam of radio waves than a motorized antenna, efficiently focusing the data beamed up and down from ground-based antennas. What is not mentioned in the presskits are the technology around optical linking between satellites. Phased arrays are old technology. The optical intersatellite links are considered by SpaceX to be proprietary and commercially sensitive.
The optical intersatellite communication may be one of the big advantages of satellite internet compared to say fiber optic transmission on the ground. Light travels almost 50% faster in the vacuum of space (~300,000 km/s) compared to through glass (~200,000 km/s). That means long-distance delays in data transmission may be much lower for Starlink internet connections than for fiber optic transmission on the ground. For an excellent discussion of data transmission rates and ion thruster propulsion, see this Real Engineering video.
One of the key regulatory requirements has been for new satellites to embody ways to remove themselves from orbit (deorbit) in order to reduce the amount of space junk circling the Earth. SpaceX plans to use the ion thrusters, used during the satellite’s lifetime to maintain and steer its orbit, to also remove itself from orbit at the end of its planned lifetime. The satellite will burn up upon re-entry into the atmosphere.
The user terminals have surprisingly been a major engineering challenge for SpaceX, perhaps because of aggressive cost constraints to keep this affordable for a global user population. Musk has repeatedly described this as a “UFO on a stick”, and images of early prototypes confirm that description. These terminals are currently going through beta testing. The FCC has already given SpaceX approval to deploy a million of these terminals.
These user terminals will also use a phased array antenna to keep track of the fast-moving satellites, rather than using a more traditional swiveling antenna. The terminal does also have motors to self-adjust for the optimal angle to view the sky.
5. An exceptional engineer and manager…
While SpaceX and Starlink were doing all this work on launches and satellites from 2015 to now, Musk was up to his eyeballs planning and executing the Tesla Model 3 and Model Y development and launches. These were not smooth roll-outs, and Musk reportedly slept on the Tesla factory floor to help problem-solve in the moment.
Musk’s executive running operations was Glynn Shotwell, President and Chief Operating Officer of SpaceX.
Shotwell was born and raised in the midwest, in Evanston, IL, the middle of three daughters of a brain surgeon and an artist. She was a straight A student in high school and loved cars. Shotwell earned a BS in Applied Math and an MS in Mechanical Engineering from Northwestern University which she attended from 1982 to 1987. So, it is no surprise that as a car-loving midwestern mechanical engineer, she planned to work in the automotive industry. She started out with Chrysler Motors in Detroit in their management training program. Her desire for a more hands-on engineering role conflicted with Chrysler’s plans, so she left. The lost opportunities for Chrysler are mind-boggling.
One of Shotwell’s professors recommended she move to Los Angeles to work for a company called Aerospace. She ended up working there for 10 years where she was involved in all areas of the company from satellite technology to launch. Shotwell moved next to a small company called Microcosm, where she encountered Musk. Musk poached her from Microcosm in 2002, when SpaceX was founded, as the eleventh employee. The rest is history.
SpaceX’s history of industry-shaking advancements like self-landing rockets, and Starlink’s unreal acceleration from a standing start to being the world’s leading satellite operator, is a testament as much to Shotwell’s executive strengths as it is to Musk’s vision.
When Starlink comes to your neighborhood, especially if you live some place without internet connectivity, make sure you thank Gwynn Shotwell as well as Musk and the rest of the SpaceX/Starlink team. These are phenomenal engineers and managers.
6. Disclosure…
I am a happy owner of about the oldest used Tesla Model S, and am also invested in Tesla the company.
Commentaires