With the surge in the number of space debris, congestion in low Earth orbit has intensified, posing a severe threat to spacecraft safety and the long-term sustainability of outer space. Traditional space debris environmental governance faces challenges such as interdisciplinary and multi-source massive data processing, while Artificial Intelligence (AI) offers a new technological pathway by virtue of its capabilities in large-scale data fusion and capturing complex correlations. This paper systematically explores the paradigm shifts in AI-driven space debris environmental governance, covering the transition from "space-based sensing and ground-based computing" to "space-based sensing and space-based computing", the shift from experience-driven to data-driven approaches, the upgrade from passive response to proactive governance, and the expansion from localized to systemic governance. On this basis, the article analyzes the applications of AI in typical scenarios such as integrated ground-space debris monitoring, optimization of spacecraft protective structures, on-orbit impact sensing and health diagnosis, low-cost removal of high-risk debris, and on-board autonomous perception and active collision avoidance. Finally, suggestions are proposed, including building a technical ecosystem, perfecting the regulatory and standard system, and advancing international cooperation, to promote the in-depth application of AI in space debris environmental governance.

The Factory in Space represents a pivotal technological paradigm enabling long-duration space missions and the construction of space infrastructure. Its core lies in achieving a closed-loop system of manufacturing, assembly, maintenance, and resource recycling under extreme space environments characterized by microgravity, high vacuum, intense radiation, and extreme temperature fluctuations. This paper systematically reviews the key technologies across the entire lifecycle of space factories, encompassing ground-based design and validation, launch ascent and deployment, on-orbit operation and expansion, maintenance and upgrade, and decommissioning and recovery. It focuses on analyzing manufacturability constraints, digital twin, lightweight structures, manufacturing process characteristics, material performance verification, on-orbit fault diagnosis, robotic maintenance strategies, and in-situ resource utilization technologies. Drawing on NASA’s Archinaut/OSAM-2 program and verification cases from China’s Space Station, the study summarizes the current technological progress and existing bottlenecks, and provides an outlook on the future development vision of space factories.