Monday, March 9, 2020

Elon Musk Keynote - Satellite 2020

Elon Musk delivered the keynote at Satellite 2020. These were my notes with some comments.

  • innovation is insufficient
  • the sign of success is not going bankrupt
  • focus on achieving something soon, if the schedule is wrong, the design is wrong
  • the bigger the rocket the more affordable the service
  • learning is inherently accessible to everyone
Musk was outspoken on how every LEO operator had gone bankrupt, and that Starlink would not go bankrupt. That any talk of an IPO was muted - it will wait for profitability. 

SpaceX sees the launch market as something like $3B/year and the overall broadband market as $30B/year. For this reason, Starlink is the next logical venture.  
(I am drawn to a speech I heard Alan Mulally talk about how Boeing made X building airplanes, the airlines made 10X operating them, and the travel industry made 100X from the journey. Mulally yearned to move Boeing up the value chain. Connexion by Boeing was one response - not bankrupt - shut down.)
Starlink is meant to round out coverage - to serve the 3-4% of customers that are hardest to reach or are out of reach of existing cellular markets. Starlink spot beams are too large to serve dense urban markets.  5G small cells are ideally suited to urban markets. Starlink is a complement to 5G, by filling the coverage areas needing long range and with few subscribers. Not downtown LA.
(AT&T 2019 wireless revenue was about $55B. Viasat 2019 services revenue was $680mm. What does the $30B represent? 5% of AT&T revenue is about $3B. Ten of those deals in the world is about $30B.)
Latency less than 20 msec.  Good enough to watch high definition video, gaming, whatever - experience does not become a drag - user does not notice speed as a bottleneck.
(The age old debate between network Vs experience. It is not important how many bps, all that matters is what the user expects and experiences.)
Initial user terminal will be self-erecting - will not require any installation support. Power it up and point it at the sky. There will be mechanical actuators. 
(not sure if hybrid fine electronically steerable array and rough mechanical pointing, or fixed beam array and fine mechanical pointing.)
Musk was a bit coy in his answer regarding interference with visible-light astronomers. Musk claimed there would be zero impact on their discovery or science. He believes that the marketplace is too focused on the launch phase, when the satellites are in lower orbits and may not be optimally aligned in attitude. He claims no one can loecate every one of the satellites - that they do disappear when fully in position. They are exploring black paint instead of white paint on the arrays. They are looking at a small sun-shade: the reflection can be blocked with a small ridge, because when the reflections are objectionable, the angles are severe.
(I worry about a sky filled with satellite dots wizzing about and obscuring the stars - I think this is a global issue. The visible-light astronomers do not have the same rights as radio astronomers. It is interesting to note how differently Iridium flares were considered).
Starlink and Starship are the two projects Musk is focused on. 

Falcon 9 was created by about 100 persons, today SpaceX employs 6000. It is much harder to manufacture than to invent - 10x or more.  Key to any startup is to focus on doing something soon.  If the schedule is wrong, the design is wrong. Simplify and focus.

Software has existed for many years. Maintaining software has a shelf life. Software needs to be re-written periodically. The same is true with rockets. Falcon 9 and Dragon are operational. The next step is to go bigger. Affordability is driven by reusability and scale. Starship is much bigger and is designed to launch one hour after landing - only the propellant is replaced. 

Musk mentions how the oceans are crossed by large container ships, each carrying hundreds of containers; and not 100s of containers with little outboard motors plying the seas. He likens the same between Falcon 9 and Starship.

Composites were abandoned because they were taking too long and costing too much. The alternative to use stainless steel turns out a better solution overall. Steel has much higher temperature tolerance, can be welded, does not need to be painted, requires less thermal shielding compared to composites and is immediately accessible. The strength from hardening and considering the cryogenic environment is equivalent to composite - the rest is all upside (less weight and headache).

Musk made some provocative statements regarding institutions of higher learning. He feels colleges only serve to show that you can answer the questions asked, but that these are not the right questions. That a degree is not evidence of achievement other than going to class and filling out the homework. He does not expect a college degree to be a requirement for working for SpaceX.
(Musk went off the deep end here. He first whines about the lack of innovation and then he finishes bashing the one marketplace that delivers more innovation than any other, and lights off more careers, and lifts up more people)

Stay Tuned!

Peter Lemme


  1. "off the deep end" is part of his social media strategy. All eyes on the weird and wonderful Musk.

  2. the universities are not a marketplace; they are a racket

  3. Problem: Career needed with huge salary, obscenely generous pension, power to hire and fire, freedoms from accountability, and no productivity metrics other than regular price increases for product. Solution? College administrator