Modern AI for Space Systems: Autonomy, Safety, and Assurance

Artificial intelligence is reshaping space systems by enabling higher levels of autonomy, efficiency, and adaptability in exploratory and operational missions. Despite significant advances in foundation models, learning-based control, and sensor fusion, deploying AI safely and reliably within real-world space environments remains a critical challenge. Studies have demonstrated the potential of AI in tasks such as mission planning, guidance, and perception, but operational constraints—like limited computational resources, latency, and harsh conditions—pose persistent obstacles to assurance, verification, and robust performance.This Research Topic aims to advance the foundations and practical deployment of modern AI in space systems while addressing autonomy, safety, and verification under real operational limitations. We seek contributions that clarify how resource-aware, reliable, and certifiable AI can be designed for spacecraft, space situational awareness, autonomous system operations, resilient platform management, and ground operations. The objective is to bridge the gap between research and deployment, fostering innovation in both technical methodologies and governance frameworks, and refining best practices through open datasets, benchmarks, and critical evaluation.The scope encompasses fundamental research, applied work, evidence from operational systems, as well as perspectives and roadmaps, and emphasizes reproducibility and operational viability. To gather further insights, we welcome articles addressing, but not limited to, the following themes:• Foundation and multimodal models for mission planning and situational awareness• Safe reinforcement learning and robust autonomous control• Multimodal perception and data fusion in space environments• Autonomous platform and system management, including autonomous Fault Detection, Isolation, and Recovery (FDIR)• Resource-aware AI: onboard algorithms for constrained platforms• Methods for assurance, verification, and adversarial robustness• Human-autonomy teaming, governance, and ethical considerations.Submissions must include an Operational Constraints & Reproducibility Statement covering: compute/power/memory/latency, evaluation design (baselines, ablations, OOD/sim-to-real), safety/robustness testing, and artefact availability (code/data or evaluation server). We particularly encourage benchmarks with clear licenses, on-orbit or field evidence, and negative results that refine best practices. Topic editors: Santosh Bhattarai, Andrea Colagrossi, Stefano Silvestrini Source: https://www.frontiersin.org/research-topics/76872/modern-ai-for-space-systems-autonomy-safety-and-assurance