Interacting with multiple robots presents significant challenges for non-experts, especially in systems requiring concurrent behaviors. This paper introduces a plan generation method for multi-robot systems using Controlled Robot Language. The framework generates planning scripts with temporal constraints from large complex instructions, incorporating contextual awareness and resolving conflicts through mutual exclusivity expressions. Subjects, objects, and action parameters are extracted from the instructions and are grounded into robotic actions. Through linguistic evaluations and real-time experiments, the framework demonstrates its robustness in language comprehension and the ability to capture and provide temporal solutions for multi-robot planning. This work contributes to the accessibility and usability of multi-robot communication for non-experts using natural language.