Climate change demands fundamentally new approaches to decision-making. Where to invest scarce adaptation resources across competing priorities, how to design carbon markets that balance efficiency with equity, and how to coordinate across jurisdictions with vastly different historical responsibilities - these are, at their core, problems of reasoning under uncertainty, constrained optimization, and multi-agent coordination.

The 2026 Planning for a Better Planet AAAI Fall Symposium</a> brings together AI planning and reasoning researchers with climate domain experts to tackle the challenge of making the planet a healthier, more sustainable place. This symposium aims to gather both the AI and climate communities to discuss opportunities for climate-motivated planning research. This includes not only AI researchers sharing new algorithms and applications but also climate experts sharing successes and challenges from their own domains, with the goal of highlighting “problems in search of solutions” to drive future climate-oriented planning research.

AI Domains

• Automated planning and scheduling
• Multiagent coordination
• Mechanism design and negotiation protocols
• Probabilistic planning under uncertainty
• Robotics task, motion, and path planning
• Operations research
• Optimization

Climate Domains

• Earth, ocean, and climate sciences
• Energy systems and decarbonization
• Land use, agriculture, and food systems
• Transportation, urban planning, and infrastructure
• Recycling, remediation, decarbonization
• Adaptation and resilience
• Climate policy, governance, and environmental justice
• Climate advocacy, policy engagement, and social impact

Foundational problem areas

• Gather data to reduce uncertainty in earth and climate systems modeling
• Optimize systems, infrastructure, and resource distribution to mitigate climate impacts
• Coordinate systems and communities for climate adaptation and resilience
• Plan for the socio-political, economic, and technical challenges of the transition to net zero
• Translate research into advocacy, engagement, and measurable social impact

Learn more about our goals

Recognizing that research to address these foundational challenges is a long-term effort, “success” in this symposium will involve achieving the following objectives, which focus on defining an agenda for climate-inspired planning research and kick-starting a community of researchers working in this area.

1. Identify opportunities for impact (“problems in search of a solution”) where automated planning, reasoning, and decision-making can support climate science, climate mitigation, climate adaptation, and climate justice. For example, planning sensor deployment to learn more effective models, planning more sustainable, equitable, or resilient infrastructure, planning for disaster recovery, and using planning to translate data and models into climate action.
2. Define research themes and directions within automated planning, reasoning, and decision-making that enable these opportunities — including not only climate-related applications of existing methods but also new areas of research inspired by these applications.
3. Cultivate a community for researchers working at the intersection of AI/robotics and climate that will launch new collaborations and research directions and facilitate climate advocacy within planning research.
4. Surface the tensions that arise at the boundary of what formal reasoning can and cannot resolve. For instance, climate mitigation offers diffuse, global benefits while imposing concentrated costs on identifiable groups. Layered on this are problems of intergenerational fairness and international equity. Fundamentally, these are normative questions. The symposium is designed to surface these tensions and probe the limits of autonomous planning by inviting climate policy researchers and social scientists into the conversation as essential partners rather than observers.
5. Foster cross-disciplinary literacy, collaboration, and opportunities:
   • For climate domain experts, this symposium will offer direct exposure to formal planning and reasoning tools that address problems at the core of climate decision-making: decision-making under uncertainty, multi-stakeholder coordination, constrained resource allocation, and risk-aware policy design.
   • For planning and reasoning experts, this symposium will offer direct exposure to the challenges and opportunities of developing and deploying AI and robotics applications in service of climate challenges, including navigating unstructured environments for robotics, multi-faceted planning, optimization, and decision support domains for climate justice, etc.

Call for Participation (submission period closed)

We invite submissions and participants from across the planning and climate domains who are interested in or have engaged in

1 climate-inspired planning research,

2 climate applications of planning methods, and

3 identifying planning/decision-making challenges within climate domains (aka "problems in search of solutions").

We invite both short and long papers at various levels of maturity, including extended abstracts that highlight more speculative or prospective work, short papers that highlight case studies or works in progress, and full papers that present more mature solutions. We also welcome submissions that identify aspirational future directions or opportunities for climate-motivated AI planning.

We specifically invite submissions from participants in climate domains that highlight challenges representing novel opportunities for collaboration with the planning research community. In particular, we welcome extended abstracts and short papers that highlight interesting planning or decision-making challenges or case studies from climate domains, and do not require these challenges or case studies to be formalized as decision or planning problems. We invite participation from across climate domains, including from climate scientists, environmental economists, policy researchers, and practitioners working on a variety of topics, including carbon pricing and market mechanisms, international climate policy coordination, national climate strategy development, adaptation investment planning, energy transition pathways, climate risk assessment, and environmental justice. We welcome submissions highlighting challenges faced by, and work with, communities across geopolitical and social contexts, especially those most impacted by climate change, including indigenous communities and those in the Global South.

Author Recommendations

Here are some recommendations designed to help you align your submission with the goals of this symposium:

1. Relevance to Planning for a Better Planet: Explicitly articulate how your work serves to bridge climate challenges with AI planning/reasoning techniques.
2. Contribution & Impact: Highlight the novelty and value of the work:
   • Planning/Application Papers: Clearly articulate what makes your planning approach or its real-world climate application novel and effective.
   • Domain/Challenge Papers: Define the climate challenge clearly and highlight why it provides an interesting challenge for the AI planning community.
3. Correctness/Soundness:
   • Planning/Application Papers: Demonstrate that your methods, theoretical arguments, and empirical evaluations are rigorous and trustworthy.
   • Domain/Challenge Papers: Ground your problem in real-world physical and ecological constraints. Be transparent about simplifying assumptions, and document any expert or stakeholder validation.
4. Clarity & Interdisciplinary Accessibility: Write for a Broad Audience. Define terms, acronyms, and domain contexts clearly so researchers outside your primary field can fully grasp your contributions.
5. Foster Cross-Disciplinary Dialogue: Frame your open questions, limitations, and future directions to invite meaningful discussion across AI and climate domains.

Submission Guidelines

• AAAI two-column, camera-ready format (per the AAAI-26 author kit; Overleaf template available here)
• Single-blind
• Submission length:
  • Extended abstracts: 2 pages + 1 page for references/appendices.
  • Short paper: 4 pages + 1 page for references/appendices.
  • Full paper: 8 pages + 1 page for references/appendices
• EasyChair submission site: https://easychair.org/conferences/?conf=fss26

To help us provide robust, vibrant, and valuable reviews and feedback for submissions across the disciplinary spectrum, we ask that at least one member of each author team volunteer to serve as a reviewer. You can volunteer by responding to this form.

Important Dates

• September 1: Submission deadline
• September 15: Notification of acceptance
• September 29: Camera-ready deadline
• October 2: Early Registration deadline
• November 5-7: Fall Symposium Series, Westin Arlington, Arlington, Virginia, USA

Registration and Attendance

Please see the AAAI 2026 Fall Symposium website for registration and attendance information.

Connect

Please join our mailing list to stay informed about opportunities related to this event:

• Join via web (Gmail address required): Visit https://groups.google.com/g/plan4planet-l/ and click “Ask to join group”
• Join via email: Send a blank email to plan4planet-l+subscribe@googlegroups.com

Follow on Bluesky: https://bsky.app/profile/plan4planet.bsky.social

Organizing Committee

• Jim Boerkoel, Harvey Mudd College
• Charles Dawson, Emerald AI
• Jeremy Frank, NASA Ames Research Center
• Alice Lépissier, Green Finance AI

Sponsors

Contact Us

• info@plan4planet.org

Confirmed Speakers

Accepted Papers

⛈️ Disaster and Hazard Adaptation

• Compound-Hazard Feasibility Gaps in Climate Adaptation Planning
  Srimonti Dutta and Akshata Kishore Moharir (WAI USA)
• Data Fragmentation as a Critical Bottleneck in Disaster Response Planning: The EVAC-AI Case Study for Flood Risk Assessment
  Gayatri Ramchandra Vaidya, Yasas Senarath, and Hemant Purohit (George Mason University)
• What Aggregate Accuracy Hides: A Failure-Mode Analysis of Language-Guided Disaster Segmentation for Mitigation Planning
  Prince Gnan Praveen Alikana, Samruddhi Pradeep Deshmukh, and Hemant Purohit (George Mason University)

🤝 Communities, Health, and Stakeholders

• Explainable and Robust Representation-Aware Planning for Climate-Health Adaptation in India
  Yajnaseni Dash (KIET Deemed to be University) and Sudhir Chandra Sarangi (All India Institute of Medical Sciences, New Delhi)
• Recentering Care in Climate Data Work as Integral for Climate AI
  Amelia Dogan (University of Washington), Mingxin Zhang (Global Energy Monitor), Cayla Key, and Lindah Kotut (University of Washington)
• From Delegates to Trustees: Aligning LLM Planners with Absent Stakeholders in Climate Decision-Making
  Zhangfei Yang (The George Washington University)

⚖️ Climate Policy, Governance, and Economics

• Can This Climate Plan Actually Be Implemented? Testing Climate Plan Executability under Institutional Constraints
  Srimonti Dutta and Akshata Kishore Moharir (WAI USA)
• Stress-Testing CBAM under Export Reallocation with RICE-M: A Multi-Agent Reinforcement Learning Climate-Economic Model with MRIO-Calibrated Trade
  Phillip Wozny (Tilburg University), Andrew Williams (Université de Montréal; Mila – Quebec Artificial Intelligence Institute), Koen Ponse (Leiden University), Kai-Hendrik Cohrs (Universitat de València), and Erman Acar (University of Amsterdam)
• Reinforcement Learning from Ecological Feedback (RLEF): A Framework for Aligning Language-Model Planning with Planetary Boundaries
  Zhangfei Yang and Aizierjiang Aiersilan (The George Washington University)

🕵️ Environmental Monitoring and Sensing

• Toward Adaptive Planning for Autonomous Glider Observations of Subsurface Marine Heatwaves
  Charlotte Z. Reed, Victor-Alexandru Darvariu (University of Oxford, Oxford Robotics Institute), Helen L. Johnson (University of Oxford, Department of Earth Sciences), and Nick Hawes (University of Oxford, Oxford Robotics Institute)
• Physics-Guided Multi-Robot Planning for Climate Sensing
  Benjamin Shaffer (University of Pennsylvania), Victoria Edwards (University of New Hampshire), and M. Ani Hsieh (University of Pennsylvania)
• Planning Wildlife Monitoring Under Ecological Uncertainty: Scenario-Based Sensor Placement for Resource-Constrained Conservation Programs
  Hannah Murray (University of Southern California), Amrita Gupta (Microsoft AI for Good Lab), Arielle W. Parsons, Justin P. Suraci (Conservation Science Partners), and Bistra Dilkina (University of Southern California)

⚡ Energy Systems and Decarbonization

• AI Planning as Solution and as Consumer: Computational Energy as a Planning Resource
  Ilche Georgievski (University of Stuttgart)
• MicrogridOS: Agent-Orchestrated Generation of 80% Microgrid Designs for Energy Access
  Vladimir Abdelnour, Nathan Johnson, and James Nelson (Arizona State University)
• Refined Thompson Learning for Adaptive Bandits: Power-Efficient Flexibility Scheduling Across Data Centers
  Zixi Chen, Yifu Ding, Ruicheng Ao, David Simchi-Levi, and Thomas Magnanti (Massachusetts Institute of Technology)

🚉 Transportation and Built Environment

• Some Lessons from Smart Transportation to Help Plan for a Better Planet
  Biplav Srivastava (AI Institute, University of South Carolina)

Tentative Schedule

Coming Soon!