Opportunity for Visiting Researchers in Agentic AI Systems for Earth Observation

Deadline: August 31, 2026 at 23:59 CET

The ESA Φ-lab CIN is now opening applications for visionary researchers and innovators in Transformative Technologies to strengthen ESA Φ-lab’s ongoing and future projects in Earth Observation (EO). Find out more and apply now.

Main Information

🖊️ Topic: Agentic AI Systems for Earth Observation

🎓 Open to: Master's students, PhD students, postdoctoral researchers, research engineers, and early-career researchers from academia or industry

📍 Location: ESA Φ-lab, ESRIN, Frascati, Italy

📅 Duration and timing: October or November, can last for 3-6 months

Background

Recent advances in foundation models have accelerated the development of Agentic AI systems capable of autonomously reasoning, planning, and interacting with external tools to accomplish complex tasks. These systems have the potential to transform the way Earth Observation (EO) data are accessed, processed, analysed, and exploited by enabling intelligent assistants that can execute scientific workflows from high-level user objectives.

In the EO domain, Agentic AI can bridge the gap between users and the rapidly growing ecosystem of satellite missions, geospatial datasets, processing frameworks, and analytical services. Beyond traditional question answering, modern AI agents are expected to perform long-horizon planning, coordinate multiple tools, adapt to changing information, leverage domain-specialized foundation models, and produce scientifically grounded results.

The ESA Φ-lab Collaborative Innovation Network (CIN) invites highly motivated visiting researchers to contribute to the next generation of Agentic AI systems for Earth Observation. The programme aims to foster collaboration between researchers in Artificial Intelligence and Earth Observation while advancing open research and innovation in autonomous, trustworthy, and scientifically grounded AI systems.

Research Challenges

Applicants are invited to contribute to one or more of the following research directions:

Primary Research Objectives

OBJ1 – Reliable Agentic AI for Earth Observation

Develop novel methodologies enabling AI agents to autonomously solve complex EO tasks through reasoning, planning, and adaptive interaction with heterogeneous EO tools, datasets, and services, leveraging modern agent architectures and interoperable AI tool ecosystems.

OBJ2 – Domain-Adaptive Agentic Intelligence

Investigate the role of agentic systems in organising, discovering, retrieving, and benchmarking geospatial foundation models, providing structured access to information on model capabilities, applicability, datasets, tasks, and performance across EO use cases, and evaluating their benefits over general-purpose models.

Secondary Research Objectives

OBJ3 – Long-Horizon Planning and Scientific Reasoning

Advance methods that enable EO agents to decompose complex scientific objectives into reliable multi-step workflows while dynamically adapting to intermediate results.

OBJ4 – Trustworthy Tool Use and Autonomous Decision Making

Develop approaches that improve tool selection, execution, verification, self-correction, uncertainty awareness, and interoperability when interacting with diverse EO resources, APIs, and AI-enabled services.

OBJ5 – Benchmarking and Evaluation of Agentic AI

Design methodologies, benchmarks, and evaluation protocols for assessing reasoning, planning, tool use, efficiency, robustness, and scientific correctness of Agentic AI systems for Earth Observation.

Who Are We Looking For?

We are looking for highly motivated Master's students, PhD students, postdoctoral researchers, research engineers, and early-career researchers from academia or industry who are interested in advancing the state of the art in Agentic AI.

Selected candidates will join the ESA Φ-lab Collaborative Innovation Network for an onsite research visit, typically lasting 3–6 months, working closely with ESA researchers and collaborators on cutting-edge AI technologies for Earth Observation.

We particularly encourage applicants interested in establishing long-term collaborations through internships, thesis projects, joint publications, or future research opportunities.

Previous experience in Earth Observation is appreciated but not required. We welcome applicants with strong Artificial Intelligence backgrounds who are interested in applying their expertise to Earth Observation challenges.

If selected, ESA will provide an invitation letter (see template here).

Desired Background

Applicants should have a background in one or more of the following fields:

• Artificial Intelligence
• Computer Science
• Machine Learning
• Earth Observation
• Remote Sensing
• Geoinformatics
• Data Science
• Applied Mathematics
• Physics or related disciplines

The ideal candidate should have experience or a strong interest in one or more of the following topics:

• Large Language Models and Foundation Models
• Agentic AI systems
• AI reasoning, planning, and autonomous decision-making
• Tool-augmented AI and AI workflow orchestration
• Multi-agent systems
• AI evaluation and benchmarking
• Scientific machine learning
• Earth Observation and geospatial AI (preferred but not required)
• Open-source software development

Experience with modern Agentic AI frameworks and ecosystems is highly desirable, including (but not limited to):

• LangGraph and LangChain-based agent architectures
• Google Agent Development Kit (ADK) and similar agent orchestration frameworks
• Microsoft AutoGen and multi-agent collaboration frameworks
• LlamaIndex and retrieval-augmented agent architectures
• Model Context Protocol (MCP) and interoperable tool-use frameworks

Strong programming skills in Python and familiarity with modern AI frameworks (e.g., PyTorch, TensorFlow) are highly desirable.

What We Offer

Visiting researchers will have the opportunity to:

• Collaborate with researchers at ESA Φ-lab on cutting-edge AI research.
• Contribute to the development of next-generation Agentic AI systems for Earth Observation.
• Work with state-of-the-art AI models, EO datasets, and computational infrastructure.
• Explore the integration of foundation models, reasoning systems, and autonomous AI agents into scientific workflows.
• Participate in open-source software development, benchmark creation, and reproducible AI research.
• Contribute to scientific publications, technical reports, and international collaborations.
• Become part of an active international research network spanning academia, industry, and the European Space Agency.

Duration and Location

The visiting research period is typically 3–6 months, although alternative durations may be considered depending on the proposed collaboration.

The research will be conducted onsite at ESA ESRIN (Frascati, Italy) within the ESA Φ-lab Collaborative Innovation Network.

How to Apply

Interested candidates are invited to submit:

• A detailed Curriculum Vitae (CV);
• A motivation letter (maximum one page) describing:
• their background and research interests;
• their motivation for joining the programme;
• how they envision contributing to Agentic AI research for Earth Observation;
• Optional supporting material, such as publications, GitHub repositories, project portfolios, or recommendation letters.

Selection Criteria

Applications will be evaluated based on:

• Academic background and technical expertise;
• Relevance of previous experience to Agentic AI and Earth Observation;
• Programming and software development skills;
• Experience with modern AI frameworks and research methodologies;
• Potential to contribute to ongoing research activities at ESA Φ-lab;
• Availability and suitability for an onsite collaborative research visit.

About the ESA Φ-lab Collaborative Innovation Network

The ESA Φ-lab Collaborative Innovation Network (CIN) brings together researchers from academia, industry, and the European Space Agency to accelerate disruptive innovation in Artificial Intelligence for Earth Observation.

The network promotes open science, interdisciplinary collaboration, and the development of trustworthy AI technologies that enable new scientific discoveries and operational capabilities.

Visiting researchers will work alongside experts in Artificial Intelligence and Earth Observation on emerging research areas including foundation models, Agentic AI, scientific reasoning, geospatial intelligence, AI-enabled data exploitation, and open-source AI ecosystems.

Please, fill this form: