PO-PDDL: Learning Symbolic POMDPs from Visual Demonstrations for Robot Planning Under Uncertainty
arXiv:2606.15654v1 Announce Type: cross Abstract: Real-world robot task planning must operate under both stochastic action execution and partial observability, yet constructing Partially Observable Markov Decision Process (POMDP) models for real robotics domains remains difficult and labor-intensive. We introduce PO-PDDL, a symbolic formulation of POMDPs that preserves the relational structure and LLM-friendly syntax of the Planning Domain Definition Language (PDDL), while explicitly modeling partial observability, stochasticity, and beliefs. Building on this formulation, we propose a demonstr
The increasing sophistication of robotic hardware and AI models necessitates more robust and reliable planning under real-world uncertainties for general-purpose robot deployment.
Improved POMDP modeling for robotics allows for safer and more autonomous robot operation in complex, uncertain environments, accelerating their commercial viability.
This research provides a more accessible and maintainable method for designing planning models for robots operating with partial observability and stochastic outcomes.
- · Humanoid Robotics Developers
- · AI Agents Researchers
- · Logistics and Automation Companies
- · Defence Industry
- · Manual Labor in Repetitive Tasks
More capable and deployable autonomous robots in diverse real-world applications.
Accelerated adoption of robotic agents in industries requiring complex planning under uncertainty.
Reduced human intervention in hazardous or tedious tasks, potentially shifting labor markets and increasing demand for robot maintenance and oversight roles.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI