PLANTOR is a novel framework that combines Large Language Models with Prolog-based knowledge management for effective multi-robot planning, enabling complex temporal task coordination.

• Two-phase knowledge generation creates reusable, robot-oriented knowledge bases
• Three-step planning process handles temporal dependencies and resource constraints
• Parallel task execution capability optimizes multi-robot workflows
• Real-world validation in assembly scenarios demonstrates practical industrial applications

This research bridges the gap between natural language understanding and robotic execution planning, offering significant advancement for manufacturing automation, construction coordination, and industrial robotics systems.

A Temporal Planning Framework for Multi-Agent Systems via LLM-Aided Knowledge Base Management