I was pleased to have Jordan Noone and Jenna Bryant on the Acquisition Talk podcast to talk about how commercial companies in space and deep tech can do business with DoD. Jordan is General Partner of Embedded Ventures and had co-founded the space launch company Relativity Space. Jenna Bryant is a General Partner and CEO of Embedded Ventures, and was an investor and teambuilder before that.
Embedded Ventures is a venture firm that pursues the next generation of space companies — those new applications that are enabled by rapidly decreasing launch costs. They recently announced a cooperative R&D agreement (CRADA) with the Space Force where they will help companies navigate customers, requirements, and the acquisition process, and on the flip side, they help SpaceWERX evaluate the technical feasibility and talent of the companies.
In the episode we discuss:
- Why companies need to think about adversarial capital early
- How government and industry can solve each others valleys of death
- The future of of digital engineering workflows
- When dual-use companies should start engaging government
- How public affairs gets in the way of customer discovery
Download the full-text transcripts.
Prescriptive vs. Performance Requirements
As commercial technology accelerates across a number of fronts, it will be difficult for government to keep up when it defines its requirements in a very prescriptive manner. Often, specifications make it into contract solicitations based on heritage architectures, decades old technologies stacks, and diminished supply chains.
For example, “space rated” electronics that are decades old have been eclipsed in quality, performance, and price by electronics in the automotive industry. Commercial electronics are “made in bulk and have been innovated upon so many times” that they surpass legacy space designs that have become “such a closed ecosystem.” This makes them viable for use in the harsh conditions of orbit.
A performance requirement, as Jordan notes, does not prescribe how the result is achieved. The electronics must stand up to a certain level of radiation, or must compute at a certain frequency. Simpler statements of outcomes would go a long way to making it easier for new entrants to compete.
Unfortunately, Jordan continues to see most of government using prescriptive approaches. I gave the example of a NASA solicitation for the Artemis program which presumed a three-stage process for landing on the moon — the way the US did it in 1969 and the way proposed by Lockheed and Boeing. With Starship, SpaceX needs fewer stages, and while it is potentially a more cost-effective solution, was non-compliant until NASA updated its solicitation.
Future of Aerospace
At Relativity Space, Jordan helped push the idea of 3D printing aerospace parts at scale, which can dramatically reduce parts counts, costs, and increase quality. 3D printing applies pretty readily where existing design constraints make traditional manufacturing difficult. For example, in order for rocket engines to withstand the extreme heat, one solution is to have internal cooling channels (regenerative cooling). This is particularly difficult to build with traditional manufacturing, but 3D printing solves it pretty easily.
This application of 3D printing, however, is what Jordan called a “linear way to think about it… taking parts that had design constraints that were directly solved by 3D printing.” Instead, Jordan asked, “What if you designed parts that rather than just this giant mechanical assembly, you eliminate the bolts. You add internal routing. You add internal cavities. You add surfaces you couldn’t have otherwise. You add complexity that would be impossible to manufacture traditionally.”
By expanding 3D printing, companies can enable a “hardware iteration loop” on the shop floor that complements the speed of software “by having all of that feedback, all of that data, taking the uncertainty of human interaction with the designs.” Critically, this allows for hardware products to undergo continuous improvement in design without upsetting traditional manufacturing flows and tooling — thus avoiding cost penalties to rapid feedback from users and operational environments.
One company that Embedded Ventures decided to help co-found is KittyCAD, which is working to automate a lot of the digital workflow associated with engineering. For example, just changing the location of a port control protocol in a server rack requires engineers to click in user interfaces, redesign bolts, brackets, redoing drawings, calling machine shops, manually adding tolerances and changing supply chain needs. Here’s Jordan’s vision for hardware:
It’s automatically running what’s called continuous integration, continuous deployment in software where you make a change, everything else compiles, and it pushes to production. That’s all automated. You build the automated infrastructure to test, to deploy, to run, to verify it’s performing well.
As all these defense-relevant technologies come to fruition, DoD and industry will have to find a middle ground where they can work together. Companies working these deep tech problems will need early customer adoption, and DoD can provide that bridge for a company’s valley of death. But Jordan warns of startups adopting a “scorched earth” policy with government. He recommended engaging early. At Relativity, they had their first DC roadshow when they were just three or four people. But relationship building is key for future success. He recommends “focusing on that early before, there’s an ask, before there’s a contract, before there’s an emergency that needs a regulatory miracle to solve.”
Thanks Jordan and Jenna!
I’d like to thank Jordan Noone and Jenna Bryant for joining me on the Acquisition Talk podcast. You can find more about Embedded Venture here. There’s a Space News article on their CRADA with the Space Force here. Follow them @TheJordanNoone and @Jenna_Bryant. Listen to Jordan on the People of Science podcast, the Orbital Mechanics podcast, and the Space Show. Listen to Jenna on Today in Space and Freight Waves.
Leave a Reply