Hoang Minh Pham

This research focuses on developing an autonomous robotic pruning system for vineyards, addressing the challenges of dynamic outdoor environments. The system integrates advanced perception and manipulation frameworks to achieve real-time modelling of branch structures and precise pruning actions. For perception, multi-view 3D reconstruction with eye-in-hand vision systems and adaptive LED lighting enhance visibility and modelling under variable conditions. Manipulation strategies include heuristic solutions for constrained pruning task optimisation, adaptive strategies for collaborative and independent dual-arm pruning, and robust visual and tactile feedback integration to ensure precision. By combining these elements, the research advances the design of scalable robotic systems for agricultural applications.