Space computing and space Artificial Intelligence (space AI) are emerging as foundational technologies for next-generation space systems and space-ground integrated networks. Space computing extends cloud-edge computing paradigms into space environments, enabling distributed processing across satellites, high-altitude platforms, space stations and ground infrastructures. Space AI further enhances these systems by enabling autonomous perception, decision-making and control for complex space missions. Together, they provide the computational and intelligent backbone for applications such as satellite constellations, space situational awareness, deep-space exploration, Earth observation, space robotics and space-air-ground integrated networks.
Recent advances in large-scale satellite constellations, inter-satellite networking and onboard intelligence have highlighted both the opportunities and challenges of space computing and space AI. Space systems must process heterogeneous, high-volume data streams from remote sensing instruments, navigation payloads, scientific sensors and communication networks under extreme constraints of power, bandwidth, latency and reliability. AI models deployed in space must be efficient, robust and adaptive, combining physics-based models with data-driven and hybrid learning approaches, and supporting online learning, federated learning and collaborative inference across distributed space-ground resources. At the same time, computation, communication and control must be tightly coordinated across space, air and ground segments to enable real-time or near-real-time services.
As space infrastructures become more autonomous and intelligent, critical issues arise in system architecture, resource orchestration, resilience, security, privacy and trust. Space AI systems must operate reliably in harsh environments, tolerate faults and radiation, and ensure safe and explainable decision-making for mission-critical tasks. Clear performance and quality metrics are needed, such as computing efficiency, inference latency, energy consumption, model accuracy, autonomy level, coordination overhead and recovery time. In addition, standardization, interoperability and governance will play a key role in enabling large-scale deployment of space computing and space AI across international and multi-vendor ecosystems.
This special issue on space computing and space AI invites contributions that address these challenges for space information networks and intelligent space systems. We welcome original research papers, surveys and position papers on architectures, platforms and algorithms for space computing and space intelligence, as well as experimental platforms, testbeds, simulations and real deployments. Interdisciplinary studies linking space systems with communications, networking, robotics, remote sensing, autonomous systems and policy are particularly encouraged.
