Technical note / 2026-08-28
A prospecting mission needs a return path
Finding a useful resource is only half a mission. Evidence has to cross a changing, delayed, and occasionally disconnected network to become a decision on Earth.
Argentora is a deep-space communications company. This note uses water-resource reconnaissance at asteroid-belt targets as a test case for the data infrastructure future missions will require. AstraNode is the relay and sensing-node program; AstraShift is the network layer that returns measurements, target assessments, and mission state to Earth.
AstraNode and AstraShift are active development programs. Current work combines deployable solar generation, battery storage, electric propulsion, optical terminals, onboard sensing, and edge compute. Trajectory, power, thermal behavior, pointing, autonomy, and link performance are coupled design constraints. Resource reconnaissance and communications are one mission system.
The network
The return path is part of the instrument.
A reconnaissance node can collect useful evidence only if it can preserve, prioritize, and forward that evidence through delayed contacts and changing geometry. AstraShift is the coordination layer Argentora is developing to evaluate those changing paths. AstraNode is the autonomous relay and sensing-node program. Each node receives, stores, routes, and forwards data while participating in the network it depends on.
Reference architecture
Treat the node as part of the network.
The useful output is a chain of custody for mission evidence. AstraNode acquires an observation, evaluates it locally, assigns priority, and preserves the original data. AstraShift determines when and where that evidence should move as contacts become available.
- 01Observe
- 02Assess
- 03Prioritize
- 04Store
- 05Select route
- 06Relay
- 07Deliver
The autonomy
Edge AI should make bounded decisions close to the hardware.
Argentora’s onboard autonomy works from versioned mission state and proposes bounded actions with confidence and expiration. Deterministic controls enforce power, thermal, pointing, navigation, communication, and safe-fallback limits. This structure reduces the raw data and operator intervention required across delayed links while preserving a record of how a decision was made.
Validation path
Prove the data path closer to Earth.
Validation begins in LEO with high-rate data delivery, interrupted contacts, store-and-forward routing, and bounded onboard decisions. Cislunar scenarios extend the same architecture across longer delays and changing geometry. The asteroid-belt reconnaissance case tests the complete system under power, distance, and contact constraints.
Contact loss, recovery, and high-rate return.
Longer delays, moving geometry, and scheduled paths.
Power-limited sensing, prioritization, and relay.
Each stage reduces uncertainty in the path from an observation at a remote target to a decision back on Earth.