The intensification of controlled environment agriculture (CEA) has revealed critical pollination challenges due to the decline of natural pollinator populations and the incompatibility of traditional bee-based pollination in enclosed systems such as greenhouses and vertical farms. This article examines the development and deployment of robotic bees as autonomous artificial pollination systems specifically designed for controlled agricultural environments. We review the mechatronic design principles, actuation mechanisms, and aerodynamic configurations that enable micro aerial vehicles to replicate pollinator flight dynamics and flower interaction behaviors. The integration of computer vision systems, artificial intelligence algorithms, and precision sensing technologies facilitates autonomous flower detection, navigation, and targeted pollen transfer through contact-based, electrostatic, and airflow-mediated mechanisms. System architectures incorporating swarm intelligence enable coordinated multi-robot pollination strategies that optimize coverage efficiency and crop yield in greenhouse tomatoes, strawberries, and vertical farming operations. Key advantages include elimination of allergen exposure, precise pollination timing, data-driven crop monitoring, and operation independence from environmental conditions that limit natural pollinators. Current limitations encompass energy density constraints affecting flight duration, scalability challenges for commercial deployment, and cost-effectiveness relative to existing manual and mechanical pollination methods. Future research directions emphasize bio-inspired design optimization, enhanced autonomy through deep learning, integration with digital twin frameworks for precision agriculture, and development of standardized protocols for robotic pollination systems in sustainable food production infrastructures.