Robotic cells are serial production systems that consist of a number of machines and a material handling robot that transfers parts between the machines. Energy consumption of a robot during its move between two machines can be decreased by carefully selecting its speed. This paper presents models and algorithms that make robot speed and robot activity sequencing decisions together for a dual-gripper robot that can handle two parts simultaneously. We consider both robot energy consumption and cycle time objectives so that a decision maker can observe the trade off between energy cost and throughput rate of the system, and select the best solution among alternatives. Computational experiments demonstrate that proposed approaches can generate efficient solutions for practical size instances, and speed control can significantly conserve energy.