The SpaceX (and Tesla) masterplan (Part 1)

(This is a followup post to (now) four earlier posts on forecasting.

  • The first in May 2015 forecast both blimp-based and dedicated building-based drone deployments (later patented by Amazon);
  • The second in October 2015 largely predicted Elon Musk’s Tesla Masterplan Part Deux by 9 months
  • The third in July 2016 among other things correctly hypothesised the use of Model X falcon wings for future possible Tesla bus designs.
  • The fourth was yesterday on SpaceX telecoms plans. This is the expanded version.
  • I try to get it mostly right but I mainly love the idle speculation).

In the mode of Elon, I propose to outline what I think SpaceX will do over the next decade, how that intersects with Tesla’s strategy of building a sustainable energy future (combined with sustainable transport) and ultimately how that fits in with making humanity a multi-planetary species.

There are a number of moving pieces to this strategy and I can’t cover them all in one blog post, but I will try to distill it down to three distinct parts.

  1. Create a space company to reduce the cost of orbital delivery
  2. Deploy an orbital ISP and get billions of customers (The X in SpaceX?)
  3. Use meta-material antennas everywhere (including Mars), creating synergies with autonomous transport systems/sustainable transport

There is far more to this, and as ever with making forecasts like this, it carries lots of provisos.

Part one

This part of the plan is already well underway.

SpaceX continues to successfully recover its first stage when the flight configuration allows it. It is increasing its stock of used-at-least-once first stages. These first stages will also double as boosters for its heavy lift rocket the Falcon Heavy when it’s due to launch later this year.

Landed first stage Falcon 9.

The cadence for launches is already impressive this year, after a couple of launch failures. All of this contributes to reducing the overall cost of delivering payloads to orbit.

Over the coming decade you can expect that the cost of delivering anything to orbit is going to fall significantly. It will become more like catching a flight — and this will create an awful lot more capacity to deploy lots more vehicles. But what demand will meet the abundant new capacity?

Part two

Here is where things get interesting over the next 10 years. Up to the end of 2016 there were 1,459 satellites in orbit around the Earth. SpaceX have a backlog of launches in the dozens.

But you have to wonder: even if SpaceX captured 100% of the entire global launch market, would the margin — even with re-use — be enough to make enough cash to continue building expensive vehicles like the Mars Transporter? I would guess no — there’s less than 100 launches per year globally.

Even if SpaceX had 100% of the launch market, and the market was growing at say 20% per year, and SpaceX was making say, $30m profit per launch (a decent margin you’d guess), then that will still only be $3bn profit — hardly enough to build a fleet for transporting 1 million people to Mars and building a colony.

And if all this launch capacity and reduction in cost was only for Mars purposes, how does SpaceX propose to pay for the Mars Transporter(s)?

So what’s the alternative?

Build a global Internet Service Provider (ISP) that requires nearly 12,000 satellites to be put into orbit. Let’s call this new ISP ‘X’ for the purposes of this piece. (it may be a spin-off company from the launch part— SpaceX).

But how do you build a global ISP? First you reduce to cost of delivering satellites to orbit. Then you:

  1. Establish an operation in Washington State (a centre for space talent) with the purpose of mass-producing satellites as efficiently as possible using techniques like 3D printing and advanced manufacturing — similar to systems developed in rocket manufacture / battery manufacture / car manufacture. Satellites essentially become standardised rather than bespoke — reducing the cost of manufacture through economies of scale.
  2. Hire engineers specialising in building beam forming / phased array technology. Seems like Field Programmable Gate Arrays (FPGA) for in-orbit configuration on the fly and custom Application Specific Integrated Circuits (ASIC) would be needed (as they are with most satellite firms I guess).
  3. Integrate new beam forming and phased array technologies into the orbital vehicles. Propose a dual system of Low Earth Orbit (LEO) and Very Low Earth Orbit (VLEO) vehicles utilising Ka-band spectrum for communications between satellites and ground gateways, and Ku-band spectrum for communications between satellites and user terminals (and later possibly Ka to terminals too).
  4. Using the enormous, growing (and somewhat redundant) Falcon 9 launch capacity begin launching the mass-produced satellites into orbit. Aim to make the satellites as small as physically possible so that as many vehicles can be delivered per launch as possible.
  5. Scale it to 12,000 satellites in VLEO and LEO. Sell it to every human and business on the Earth (but do so gradually as you build your capacity). Charge high prices to start with and then lower prices as you scale. Also sell backhaul capacity via the LEO network — by using lasers in a vacuum to communicate between satellites in orbit you can radically reduce latency — arguably even in comparison to current land-based fibre optic cables.
  6. Build gateways (ie ways to direct internet from a ground station to the satellite network and vice versa) on multiple continents. Build the gateways close to internet backbones for maximum reduction in latency. You might only need a few gateways per continent.

Part three

But what does the product look like and how do you go market?

In the style of how Tesla launched the Model S, Model X, Model 3 and the Tesla Powerwall / Tesla Solar Roof, X will launch a product to sell access to the X network. It will be presented by Musk on stage as a huge leap forward in broadband communications. It will also be presented as simple, well designed and integrated.

What form will this come in?

  1. Within 5 years all Tesla cars will come with an optional extra. An integrated connection to the X satellite broadband network. These will be integrated into the roof (or trunk/bonnet — since electric cars have more available surface area than fossil cars) and come in the form of a non-mechanical meta-material antenna. Cars will be able to be in constant communication with the network, delivering fast broadband to every car in the fleet (the antenna can keep in constant communication with the network even when moving at speed). And here’s an important point: if Tesla want to have millions of autonomous vehicles on the road in the future they will need fast data solutions for their vehicles. It seems debatable whether the existing cellular infrastructure could handle the load.
  2. Within 5 years, Tesla will launch optional extras for its solar home product. It will be an integrated X “tile” within a solar roof. The tile will have a permanent connection to the X network. Essentially every solar roof could both be collecting solar energy while also acting as your internet access. Of course it will be sold stand-alone too.
  3. X will also sell the meta material panels across industry once the orbital network has expanded. This will include aircraft (no moving parts and flush with the fuselage, saving fuel), ships, government, military etc. This will also include backhaul selling to other networks.

If you think I’m crazy, take five minutes and take a look at the video below from January 2016.

Last year Toyota invested in the Redmond Washington-based (yes!) and five year old Kymeta (who include Bill Gates as one of their investors too).

The X network will look broadly similar, but unlike Kymeta, X will not only sell terminals — they will be building the entire ISP solution — end-to-end. And of course X will not just be limited to cars — it will apply to any location, moving or static.

Here’s some key slides from Kymeta’s presentation:

Indeed you could argue that X should buy Kymeta. Kymeta boast that their system could handle 1Tb of data per month per vehicle (seemingly using the Intelsat network).

If you believe that Tesla/SpaceX are not going to pursue this type of technology then I would argue that this is underestimating Elon Musk.

Towards the end of the talk Kymeta founder Nathan Kundtz even appears to have a dig at a possible SpaceX network.

With 50 [Intelsat] satellites each the size of a bus, and new satellites going into space all the time, this is not some network that will be ready in 2025, this is the network of now.

And let’s look at who SpaceX is hiring in Redmond right now:

Once you’ve built this system, and hopefully overcome regulatory hurdles; got your first 100,000 customers; built the largest ISP in orbit over time; integrated all of your vehicles to the network — then you can use that technology and kn0w-how and scale to deploy the same system to Mars.

You would do this in advance of the colony being established. It makes more sense to pre-position a communications network at Mars than it does to build it in tandem or after astronauts have already landed there.

In summary:

  1. Sometime this year or early next year Musk will announce the plan for the “X” network. He will outline a broad schedule for launches and coverage areas and include high-level pricing.
  2. SpaceX will seek permission for various bands, and seek to overcome any regulatory hurdles. It will meet resistance from incumbents.
  3. Musk will later announce the launch of Tesla vehicles with integrated antennas to access the orbital network. Tesla will launch home products for broadband and sell X via its Tesla showroom network globally. These products will integrate will with solar roof/solar storage products.

I’m happy to bet a whole €100 that this will, in large part, be attempted. Whether or not it’s successful is another thing.

And if I am right, Elon Musk owes me a few pints of Guinness 😂.

(Disclosure: I’m a *very* small shareholder in Tesla. I’m the founder over at Vizlegal (in Ireland!) where we’re building a global API for law — a sorely needed thing if you want autonomous machines (and even a Musk Mars colony needs laws too!). I’m on Twitter if you have any questions)