Shaking hands as a gesture of greeting is a common social practice observed in numerous cultures. This behavior is acquired during the early years of childhood. We constructed a humanoid robot with the objective of greeting children and their parents at a hospital. During the robot’s initial deployment, children frequently extended their hands to grasp the robot’s hand in order to initiate a handshake. Nevertheless, social robots, including our own, require preparation for such activities. To address this challenge, we implemented an interactive handshaking protocol for our robot, which employs tactile sensors embedded in the right hand to adapt and detect the optimal timing for each phase of the handshake. In a user study, the adaptive handshake was compared with two other protocols, one with constant timing and the second with passive, compliant control, which expected the users to lead the activity. The results of statistical tests revealed that the participants expressed a preference for the adaptive robot, perceiving its behavior as more natural, pleasant, and reactive. In conclusion, it was found that the implementation of handshaking procedures on social robots with arms can facilitate more natural interaction and deepen trust. Furthermore, insights are provided regarding prospective enhancements to the platform, including integrating neuromorphic sensing technologies for sensorimotor tactile control.