To interact with humans, robots will possess a software architecture much more complete than current robots and be equipped with new functionalities. The purpose of this chapter is to introduce some necessary elements to build companion robots that interact physically with humans and particularly for the exchange of object tasks. To obtain soft motion acceptable by humans, we use trajectories represented by cubic functions of time that allow mastering and limiting velocity, acceleration and jerk of the robot in the vicinity of the humans. During a handover task and to adapt its trajectory to the human behavior, the robot must adjust the time motion law and the path of the trajectory in real time. The necessity of real time planning is illustrated by the task of exchanging an object and in particular by the planning of double grasps. The robot has to choose dynamically a consistent grasp that enables both robot and human to hold simultaneously the exchanged object. Then, we present a robotic control system endowed with attentional models and mechanisms suitable for balancing the trade-off between safe human-robot interaction (HRI) and effective task execution. In particular, these mechanisms allow the robot to increase or decrease the degree of attention toward relevant activities modulating the frequency of the monitoring rate and the speed associated to the robot movements. In this attentional framework, we consider pick-and-place and give-and-receive attentional behaviors. To assess the system performances we introduce suitable evaluation criteria taking into account safety, reliability, efficiency, and effectiveness.