Multi-Agent Modal Computing: Synergy, Scalability and Satellite-Earth Collaboration

Dear Colleagues,

The explosive growth of multi-modal data from satellite constellations, IoT sensors, and social media demands new computing paradigms that can jointly optimize perception, reasoning, and actuation across heterogeneous agents. This Special Issue focuses on Multi-Agent Modal Computing (MAMC)—a paradigm where multiple autonomous agents, each specialized in a specific modality (vision, language, tabular, sensor, spectrum, etc.), collaborate to achieve global objectives under resource-constrained, dynamic, and distributed environments such as satellite swarms and the edge–cloud continuum.

This Special Issue will (a) consolidate emerging theories that unify agent coordination, cross-modal representation learning, and distributed inference; (b) cover system-level innovations enabling on-orbit federated learning, inter-satellite communication-efficient model fusion, and real-time Earth-observation analytics; and (c) highlight practical deployments in disaster response, smart cities, and climate monitoring. By bridging multi-agent systems, modal learning, and satellite computing communities, this collection aims to complement existing research that either treats modality fusion in a single-agent setting or addresses satellite AI without multi-agent synergy.

We solicit original contributions on, but not limited to, the following:

• Agent architectures for multi-modal reasoning under partial observability.
• Multi-modal learning in knowledge graph-based, social network, and satellite contexts.
• Multi-modal learning in NLP (e.g., text mining, knowledge graphs, etc.).
• Multi-modal learning in CV (e.g., object detection, super-resolution, video–text retrieval, satellite-related applications, video tracking, etc).
• Communication-efficient consensus algorithms for modal model fusion.
• On-board/off-board co-design for large multimodal models in satellites.
• Explainable AI in multi-modal graph mining and knowledge discovery.
• Multi model neuro-symbolic learning.
• Benchmarks, datasets, and simulation platforms for MAMC.
• Security, privacy, and fairness in multi-agent modal ecosystems.
• Continual and federated multi-modal learning across heterogeneous agents.
• Energy-aware model compression and quantization for on-orbit multi-modal inference.
• Cross-agent transfer learning and meta-learning under modality imbalance.
• Human-in-the-loop collaboration between ground operators and satellite swarms.
• Causal inference and counterfactual reasoning in multi-agent multimodal systems.
• Uncertainty quantification and calibration for distributed multimodal perception.
• Real-time multimodal anomaly detection and early-warning on edge satellites.
• Graph-based neural architecture search for modality-specific agents.
• Privacy-preserving multi-modal analytics using secure multi-party computation or homomorphic encryption.
• Ethical governance and policy frameworks for autonomous satellite agent constellations.

Dr. Qian Li
Dr. Yunfei Long
Dr. Cheng Ji
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. AI is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• multi-agent systems
• multimodal learning
• satellite computing
• knowledge graph
• edge AI
• federated learning
• remote sensing
• distributed inference