🤖 AI Summary
This work addresses the challenge of efficiently translating mission intent into structured, machine-readable execution outcomes on in-orbit satellites under stringent communication and power constraints. The authors present a hardware-in-the-loop intelligent agent system deployed on a commercial ARM heterogeneous edge SoC, which orchestrates a local large language model and a YOLO-style oriented object detection module via FastAPI within a browser-based workspace to enable task orchestration and real-time state feedback. The study establishes, for the first time, an observable boundary for satellite edge intelligent agent orchestration, supports FAIR1M metadata-compliant output, and provides a fully reproducible experimental package. Empirical results demonstrate 100% success across 20 trials for two FAIR1M workloads, achieving end-to-end latencies of 29.35 s and 60.94 s, with the detection module consuming less than 3% of total runtime, average CPU utilization at 20.6%, and NPU utilization reaching 100%.
📝 Abstract
Onboard satellite intelligence requires a task layer that translates mission intent into local tool calls, exposes execution state, and returns machine-consumable artifacts under communication and power constraints. We present SAT-Edge-Agent, a hardware-in-the-loop (HIL) edge-agent system deployed on a commercial off-the-shelf ARM-based heterogeneous edge system-on-chip. A browser workspace and FastAPI agent coordinate a local OpenAI-compatible language service with a project-internal YOLO-style oriented-object-detection endpoint that returns FAIR1M metadata-backed structured results. Two fixed FAIR1M workloads, one single-image and one serial two-image request, were repeated 20 times each and completed 20/20 attempts. Mean Full-Agent latency was 29.353 s and 60.937 s, with empirical P95 values of 31.166 s and 66.882 s. Mean detector time was 861.386 ms and 1510.920 ms, only 2.93% and 2.48% of the corresponding Full-Agent means. Profiling indicates that most visible latency occurs outside detector execution. Mean CPU utilization was 20.761% and 20.482%. A 200-ms NPU-load field averaged 100% for both workloads, but it represents a shared-accelerator software field rather than detector-only occupancy or calibrated utilization. The public evidence package provides sanitized request-level records, redacted JSON, normalized SSE examples, and scripts reproducing the reported statistics. These results establish a reproducible HIL boundary for observable satellite edge-agent orchestration, but do not establish detector accuracy, a new geolocation method, calibrated energy efficiency, or flight readiness.