1. The toolbox decides the workshop

Andy Weir’s The Martian is fun because Mark Watney does not survive with one tool. He survives because he has a habitat, a rover, a chemistry kit, and the willingness to combine them in ways their designers never intended. The second keeper platform is the same. Its value is not the bus. Its value is the set of attachments that turn the bus into a place where work happens.

Entry 177 sketched the second platform as two 500 kg-class keeper nodes with standardized bays and a robotic arm. This entry asks what attachments make that platform worth building.

2. Attachment one: high-resolution earth observation

The lowest-risk, highest-demand attachment is a camera. Not a flagship science instrument, but a commercial-grade optical or hyperspectral imager that can be pointed, tasked, and downloaded on demand. Customers already pay for Earth imagery. A keeper platform that hosts a camera and sells tasking time has a familiar revenue model.

The platform’s value here is not the camera itself but the infrastructure around it: power, pointing, thermal control, data downlink, and the ability to swap the camera when a better one becomes available.

3. Attachment two: data storage and compute

A 1U storage box is boring. A multi-terabyte, radiation-tolerant storage node with on-orbit compute is more interesting. The desktop objective assumes that someone wants to put a computer in orbit and keep it there. The second platform can test that assumption with a smaller, customer-facing compute attachment.

Use cases: edge processing for Earth observation, secure data vaulting, latency-sensitive applications that want to be off-planet for jurisdictional or latency reasons. Entry 54 already wandered about the “above cloud” appeal.

4. Attachment three: additive manufacturing

A small 3D printer that can produce polymer or metal parts on demand is the attachment that turns the platform from a hostel into a workshop. Customers upload a part file, the platform prints it, and a visiting tug either picks it up or installs it on another satellite.

This is the orbital machine shop idea from earlier wonderings, but at a scale the second platform can actually host. The printer does not need to be industrial grade. It needs to prove that on-demand parts are possible and that someone will pay for them.

5. Attachment four: servicing toolkit

A robotic arm with interchangeable end effectors — gripper, cutter, drill, inspection camera — is the attachment that makes the platform a keeper in the original sense. It can inspect visiting satellites, remove launch restraints, apply thermal patches, or prepare a spent stage for reuse.

This is the riskiest attachment because it requires proximity operations and trust. But it is also the one most aligned with the desktop objective: a general-purpose orbital workstation.

6. Attachment five: hosted payload slots

The simplest attachment is no attachment at all. The platform sells empty slots with power and data. Customers bring their own payloads. This is the hosted payload business model that already exists, but the second platform makes it more flexible by offering standardized interfaces and shorter integration timelines.

This is the attachment that keeps the lights on while the fancier ones are being proven.

7. What earns the platform

A platform with only one of these attachments is a niche service. A platform with all five is a credible prototype of the desktop. The second keeper platform should launch with at least three: Earth observation, hosted payload slots, and either compute or manufacturing. Servicing can follow once the platform has flight heritage.

What this changes

  • The second keeper platform is justified by its attachments, not by its existence.
  • The five candidate attachments are Earth observation, data/compute, additive manufacturing, servicing toolkit, and hosted payload slots.
  • A credible second platform should carry at least three attachments at launch.
  • The next leisure entry can estimate the economics of these attachments.