Advancements in Autonomous Planning and Decision Making

The field of autonomous planning and decision making is witnessing significant advancements with the integration of symbolic and neural approaches. Researchers are exploring ways to combine the strengths of both paradigms to create more robust and reliable systems. A key direction is the development of neuro-symbolic frameworks that can seamlessly integrate symbolic planning with neural networks, enabling more efficient and interpretable decision making. These frameworks are being applied to various domains, including autonomous UAVs, multi-agent systems, and robotic manipulation. Notable papers in this area include: Scaling Multi-Agent Epistemic Planning through GNN-Derived Heuristics, which proposes a novel approach to multi-agent epistemic planning using graph neural networks. LOOP: A Plug-and-Play Neuro-Symbolic Framework for Enhancing Planning in Autonomous Systems, which develops a neuro-symbolic planning framework that treats planning as an iterative conversation between neural and symbolic components. CausalPlan: Empowering Efficient LLM Multi-Agent Collaboration Through Causality-Driven Planning, which integrates explicit structural causal reasoning into the planning process to improve collaboration in multi-agent systems.

Sources

Integrating Symbolic RL Planning into a BDI-based Autonomous UAV Framework: System Integration and SIL Validation

Scaling Multi-Agent Epistemic Planning through GNN-Derived Heuristics

LOOP: A Plug-and-Play Neuro-Symbolic Framework for Enhancing Planning in Autonomous Systems

CausalPlan: Empowering Efficient LLM Multi-Agent Collaboration Through Causality-Driven Planning

Toward Deployable Multi-Robot Collaboration via a Symbolically-Guided Decision Transformer

Active Learning for Neurosymbolic Program Synthesis

Built with on top of