UNO Push

UNO Push, and related research on robotic pushing, focuses on developing robust and efficient methods for robots to manipulate objects nonprehensively, primarily through pushing actions. Current research emphasizes the use of model predictive control (MPC) and deep reinforcement learning (DRL) to address challenges like contact uncertainty, object dynamics, and cluttered environments, often incorporating non-parametric learning for system modeling. These advancements are significant for improving robotic manipulation capabilities in various applications, such as warehouse automation and disaster response, where precise and adaptable object handling is crucial.