Abstract
The work done in this thesis is about the interactions between human and humanoid robot HRP-2Kai as co-workers in the industrial scenarios The research topics in the thesis are divided into two categories. In the context of non-physical human-robot interactions, the studies conducted in the 1st part of this thesis are mostly motivated by social interactions between human and humanoid robot co-workers, which deal with the implicit behavioural and cognitive aspects of interactions. While in the context of physical human-robot interactions, the 2nd part of this thesis is motivated by the physical manipulations during object handover between human and humanoid robot co-workers in close proximity using humanoid robot whole-body control framework and locomotion.We designed a paradigm and a repetitive task inspired by the industrial Pick-n-Place movement task, in first HRI study, we examine the effect of motor contagions induced in participants during (we call it on-line contagions) and after (off-line contagions) the observation of the same movements performed by a human, or a humanoid robot co-worker.The results from this study have suggested that off-line contagions affects participant's movement velocity while on-line contagions affect their movement frequency. Interestingly, our findings suggest that the nature of the co-worker, (human or a robot), tend to influence the off-line contagions significantly more than the on-line contagions.Under the same paradigm and repetitive industrial task, we systematically varied the robot behaviour and observed whether and how the performance of a human participant is affected by the presence of the humanoid robot. We also investigated the effect of physical form of humanoid robot co-worker where the torso and head were covered, and only the moving arm was visible to the human participants. Later, we compared these behaviours with a human co-worker and examined how the observed behavioural effects scale with experience of robots.Our results show that the human and humanoid robot co-workers have been able to affect the performance frequencies of the participants, while their task accuracy remained undisturbed and unaffected. However, with the robot co-worker, this is true only when the robot head and torso were visible, and a robot made biological movements.Next, in pHRI study, we designed an intuitive bi-directional object handover routine between human and biped humanoid robot co-worker using whole-body control and locomotion, we designed models to predict and estimate the handover position in advance along with estimating the grasp configuration of an object and active human hand during handover trials. We also designed a model to minimize the interaction forces during the handover of an unknown mass object along with the timing of the object handover routine.We mainly focused on three important key features during handover, and we answered the following questions, ---when (timing), where (position in space), how (orientation and interaction forces) of the handover.we present a generalized handover controller, where both human and the robot is capable of selecting either of their hand to handover and exchange the object. Furthermore, by utilizing a whole-body control configuration, our handover controller is able to allow the robot to use both hands simultaneously during the object handover. Depending upon the shape and size of the object that needs to be transferred.Finally, we explored the full capabilities of a biped humanoid robot and added a scenario where the robot needs to proactively take few steps in order to handover or exchange the object between its human co-worker. We have tested this scenario on real humanoid robot HRP-2Kai during both when human-robot dyad uses either single or both hands simultaneously.