Image
This shows Rocket Lab’s Mars Sample Return (MSR) architecture, a look at an extremely ambitious commercial planetary science mission.

In my other blog posts I’ve done analysis on the commercial satellite launch market based on data and an understanding of economics and the industry. Now I’ll foray into the famously difficult to predict relm of government/politics/bureaucracy.

A more complete proof of the previous blog post

Peter beck MSR article: https://spacenews.com/bringing-mars-to-earth-solving-mars-sample-return/

Jared Isaacman thoughts: https://x.com/SciGuySpace/status/1915800015781429444

Notes from rereading NASA’s Fucked:

  • Get more (& more concrete) examples of cost plus contracting being a negative
  • Find counterexamples of cost plus being a positive
  • Artemis is an example of NASA being horrible at systems engineering
  • Why was it good at systems engineering during Apollo? What was the management structure and how is it different now?
  • Dispersed centers make for more senators fighting on your behalf, and perverting your architecture
  • Centralized decision making allowed for more coherent architecture and design choices
  • Find counterexamples where government is more efficient than private industry (Probably not in the limit)
  • Find correlaries in history of a transition from government architectures to buying rides on commercial architectures

More notes:

  • First principles proof for why the end state of NASA is buying commercial capabilities
  • What are NASA’s responsibilities in the limit? (R&D? Funding early development? Experimental programs? Human exploration? Why even explore with humans?)
    • Essentially, do what is useful and private companies won’t do (Find correlaries)
  • This is a path-dependent process. Whether NASA has naming rights on the Mars colony is path dependent, can you be useful to SpaceX?
    • Not driving for the end state leads to a diminished role for NASA in the limit
    • Being proactive accelerates the beautiful future
  • Consider end-to-end space companies and they’re role in providing the full mission architecture to NASA (NASA sets goal, Rocket Lab builds the lander, orbiter, return capsule, etc.)

https://arstechnica.com/space/2025/08/as-the-end-of-the-iss-nears-nasa-shakes-up-program-for-commercial-replacements/

Old NASA CLD model:
Companies supply preliminary design studies (and maybe some hardware if they want) and then choose a contractor or two to build out a station.
Phase 2 would have been firm-fixed price contracts for certification of designs on the ground without hardware in space and “services”.

New NASA CLD model:
Multiple participants receive Space Act Agreements.
Formula design, acceptance, and certification will come in a follow-on certification phase, well after the Space Act Agreements phase.
Certification happens after they fly.
25% of milestone funding withheld until a successful in-space crewed demo. Minimum capability NASA is seeking is 4 crew for 1 month.

Benefits of this model:

  • Companies incentivized to build hardware and demonstrate their capabilities to continue in the competition.
  • Companies can build their own business models and architectures and get NASAs business in a broader sense than a strict design contract (Apollo model was like this too?).

The space companies that will be successful in the future will self-fund their own development programs. They will then sell this capability to NASA and other commercial customers, building a product that is useful to customers before the customers explicitly request it. Eg. SpaceX, Rocket Lab, Vast, Firefly (Lunar reconnaissance), etc.

Anduril too https://x.com/anduriltech/status/1953163828671463557