InstructMPC integrates Large Language Models with traditional Model Predictive Control to enable context-aware control systems that can adapt to real-time instructions.

• Combines human expertise and LLM intelligence to enhance predictive capabilities
• Addresses MPC's limitations in handling context-specific conditions
• Creates more responsive and adaptable control systems for complex environments
• Applicable across aerospace, energy management, and autonomous systems

This innovation matters for engineering because it bridges the gap between static control algorithms and dynamic real-world conditions, enabling more intelligent automation that can interpret and adapt to human instructions on the fly.

InstructMPC: A Human-LLM-in-the-Loop Framework for Context-Aware Control