With the ever-evolving landscape of manufacturing and industrial innovation, there is a growing emphasis on leveraging technology to enhance productivity, ensure safety, and address socio-economic challenges. Human-robot collaboration (HRC) systems, particularly those incorporating gestural interaction, are advancing to improve usability and operational efficiency. This study proposes integrating a high-level gesture recognition model into a digital twin of a manufacturing environment. The gesture recognition model identifies and interprets human movements in real-time, enabling intuitive and seamless interaction between humans and robotic systems. A lightweight computer vision framework was employed to enhance the model’s accuracy and responsiveness, ensuring robust detection and tracking of gestures. The development and implementation of this digital twin facilitate the simulation of various conditions under which robots operate, allowing for realistic testing and optimization of gesture-based interactions. An experiment with the digital twin demonstrated a high recognition accuracy for all gestures, with precision and recall values consistently above 90% and a latency of less than 100ms. By integrating the digital twin with a physical robotic system, this framework can ensure accurate and responsive gesture control, strengthening the effectiveness and flexibility of HRC in industrial settings.