Venue 5 of 6 · Aerospace systems · Conference

IEEE Aerospace Conference

An annual conference in Big Sky, Montana, with a heavy NASA, JPL and industry presence. It rewards systems papers that show something working end to end, and it's the right stage for the decentralised auction and the ESP32 testbed.

2027 deadlines passedTarget: 2028 editionFacts checked 11 Oct 2026
2027 cycle closedNeeds H4 + testbed

The 2027 cycle is closed. Abstracts were due 1 Jul 2026 and full papers 2 Oct 2026, for the conference on 6–13 Mar 2027. Target 2028, whose abstract deadline is likely around early July 2027 if the pattern holds. That date is inferred, not published.

Why it suits a later paper: this audience wants hardware and mission relevance. By mid-2027 the Phase 4 decentralised auction and the ESP32 testbed can be the headline, which is a paper no other venue on this list showcases as well.

Venue facts

ThemeAerospace systems and technology across many tracks. Relevant: autonomy, mission design and operations, space-based observatories, and avionics/computing for space.
FormatAbstract first, then full paper, reviewed before final submission. Page limits and template: check the 2028 call; not verified here.
2027 dates (passed)Abstract 1 Jul 2026 · full paper 2 Oct 2026 · reviews returned 9 Nov 2026 · final paper 8 Jan 2027 · conference 6–13 Mar 2027.
2028 datesNot yet published. Expect a similar cycle (abstract ~July 2027).
Sourcesaeroconf.org · AIAA event listing (2027)

Tailored abstract

Decentralised Hypothesis-Driven Tasking for Small-Satellite Observatories: Simulation and Hardware-in-the-Loop Evaluation

Constellations of small satellites carrying heterogeneous instruments could follow up fast astronomical transients without waiting for ground operators, but only if spacecraft can coordinate which of them takes which measurement while their inter-satellite links are intermittent and their beliefs out of date. We describe a decentralised tasking architecture in which each spacecraft maintains a posterior over competing physical explanations of an event, plus an explicit unmodeled alternative, and bids the marginal information value of its candidate measurements in an auction that tolerates stale beliefs. Spacecraft exchange per-class log-evidence and committed measurements, never raw data, and merge beliefs exactly by measurement identifier. We evaluate the architecture in a simulated low-Earth-orbit constellation propagated from real two-line elements while thinning contact schedules from full to sparse connectivity, and we replicate the communication and decision layer on ESP32 microcontrollers linked by ESP-NOW and gated by the simulated contact schedule. Result sentences pending: retained fraction of centralised-oracle performance vs. contact fraction (H4), and on-device compute per decision.

Paper plan

  1. Introduction and mission concept~1.5 ppConcept of operations: alert uplink, autonomous campaign, downlink. What the existing OrbitalNet CNP stack does today and what it lacks.
  2. System architecture~2 ppOrbital layer, feasibility, belief, utility, auction, belief exchange (Fig. 2 pipeline). Message formats and sizes.
  3. Decision layer~1.5 ppThe utility, compressed. The stale-belief auction in detail, and the exact merge with the duplicate-ID test.
  4. Simulation results~2.5 ppThe communication sweep and node-loss scenarios (Phase 4): P vs. O vs. independent greedy.
  5. Hardware-in-the-loop testbed~2.5 ppTable 4 of the manuscript, made real: wiring, firmware, measured compute and memory, message loss, sim-vs-hardware agreement. State what it does not show (radiation, RF link budgets, ADCS).
  6. Path to flight~1 ppShadow-mode operation alongside a real alert stream, a hosted-payload or rideshare demonstration, and the TRL steps needed.

Before you submit

← RASTI · All venues · Next: AIAA JAIS →