ARIE 2026: Frontier Session on Autonomy, Reasoning, and Intelligence Beyond Earth
ARIE 2026 will be held in the Cross-AI Symposium 2026.
Introduction
The next generation of space exploration, spanning autonomous spacecraft, lunar infrastructure, planetary robotics, and deep-space missions, demands a fundamental shift from human-in-the-loop operation toward AI-driven mission intelligence. As missions extend beyond Earth, reliance on ground-based computation, high-bandwidth communications, and real-time human control becomes increasingly constrained.
Space systems must operate with limited computation, energy, memory, storage, and bandwidth while facing radiation, sensor degradation, communication delays, intermittent connectivity, and uncertain environments. These conditions challenge predefined autonomy and require AI/ML capable of long-horizon reasoning, learning from sparse and distributed data, adaptive resource management, multi-agent coordination, and resilient decision-making with limited terrestrial intervention.
Autonomy, Reasoning, and Intelligence Beyond Earth (ARIE) explores the algorithmic foundations, architectures, and learning methodologies for this emerging paradigm across four interconnected directions: agentic AI and autonomous mission intelligence; resource-efficient and onboard edge AI; trustworthy, secure, and resilient intelligence; and collective, swarm, and embodied intelligence.
The workshop adopts a cross-disciplinary perspective spanning AI/ML, robotics, computer vision, aerospace engineering, embedded computing, communications and networking, information theory, applied physics, and space science. Its methods may also benefit extreme terrestrial environments where resources, connectivity, or human access are similarly constrained. ARIE brings together researchers and practitioners to advance intelligent, adaptive, secure, and resilient systems beyond Earth.
Topics
Agentic AI & Autonomous Mission Intelligence
- Mission-Level Reasoners: Agentic AI, long-horizon planning, and real-time mission replanning without ground control.
- Uncertainty & Latency Mitigation: Decision-making models optimized for multiminute feedback loops and degraded inputs.
- Goal-Directed Autonomy: Resource-aware, self-directed scheduling, execution, and in-situ autonomous science tracking.
Collective, Swarm, & Embodied Intelligence
- Multi-Agent Coordination: Distributed reinforcement learning, collective swarm intelligence, and heterogeneous fleet management (orbiters, landers, rovers).
- Embodied Space AI: Learning-based perception, active vision, slip compensation, and terrain navigation in unmapped environments.
- Sim-to-Real Adaptation: Physics-informed neural networks (PINNs) and domain adaptation frameworks for validating AI models prior to deployment.
Trustworthy, Secure, & Resilient Space AI
- Inspection Intelligence: Vision and learning models for autonomous robotic/drone-based structural defect detection under extreme environmental stress.
- Self-Healing Models: Real-time anomaly detection, self-monitoring, safety-constrained learning, and automated fault recovery.
- Space AI Security: Algorithmic protection against model hallucination, adversarial manipulation, and in-situ data poisoning.
Resource-Efficient & Onboard Edge AI
- Hardware-Aware Architectures: Sparse, heavily quantized, neuromorphic, and radiation-tolerant neural network deployment.
- Distributed Onboard Learning: Decentralized, collaborative model updates via onboard federated learning without raw data exchange.
- Cognitive Compression: Semantic, task-aware, and generative data prioritization to maximize constrained orbital bandwidth.
Important Dates
See Important Dates for all paper and abstract submission deadlines.
Submission Guidelines
Submissions are made through the AirBalloon conference management system. Full and Short Papers are reviewed double-blind.
Before submitting, read the Author Guidelines, the Submission Guidelines, the Publication and Media Hosting guidelines, and all policies, then complete the Checklist for Initial Submission. Submissions that do not follow the guidelines and policies are desk-rejected under the Desk Rejection Policy.
To serve on the workshop program committee or as a reviewer, please apply online. Applicants may select multiple workshops relevant to their expertise. Approved applicants may be invited to review papers. Those providing high-quality reviews will be appointed to the Program Committee, and others will be acknowledged as external reviewers.