Order picking is the operation that consists in retrieving, from the storage locations, the parts needed for the assembly of final products. Making a ready-to-delivery collection or kit of parts is called Kitting. To improve this operation's performance, flexible robotic systems can greatly help industrials. Indeed, the technological advances that have been achieved in the recent years, in robotics and artificial intelligence make it possible to deal with a large range of items to be picked despite some remaining constraints related to components diversity and parts characteristics. In this paper, we study a robotic kitting system running with a robot arm mounted on a rail system and traveling along a narrow-aisle to pick parts. Through a modeling of elementary kitting operations that the robot performs (pick and place, travel, tool changing, etc.), we aim at evaluating the performance of the robotic kitting system in terms of cycle times. This study conducted with a manufacturer in the context of an ongoing project on automation of kitting operations, can help him to assess the robotic area performance in a given configuration (layout, picking policy, etc.).