Rock/Paper/cissors (RPS)
Rock Paper Cissors is a complex and multimodal task that requires planning decision-making and movements synchronization, and induces emotion in case of success or failure of one partner (human and even robot) and possibly speech exchange. In this experiment a humanoid robot Pepper will be used as the master in the game against the human partner. In order to induce in the human partner affect and stress, the robot will be programmed to i) let him/her win at each time or lose at each time based on physical (data gloves, robot camera) and neuronal (electroencephalography) movement preparation, ii) to surprise him/her by ii.a) non-conventional gestures in the game or ii.b) post gesture change to win or ii.c) incoherent verbal or display results. One of the three different robot behaviors (neutral, kind or mocking attitude) will be presented to each participant. Behavior and affect of the human will be studied using sensors and questionnaires.
Rope jumping (RJ)
RJ is a rhythmic interaction with contact through a flexible object. RJ is a complex task requiring synchronization of robot and human movements through a flexible mechanical coupling (the rope) between human and robot. This task will be conducted to design and evaluate a specific bio-inspired controller to enable the robot to learn to synchronize its movements (trajectory of hand effector) to those of humans in the rotation directions. The robot controller will be based on central pattern generators (CPG) that are specific non-linear oscillating neurons [Jouaiti et al. 2019] inspired by a part of the human motor nervous system, located in the spinal cord, in charge of the control of rhythmic movements. In order to ensure adaptability, so the robot can learn to synchronize its movements to those of humans, a learning mechanism will be implemented in the CPG networks considering the forces measured at the robot’s wrist (cartesian forces) or in other joints (articular torques). Different kinds of CPG architectures will be compared. Furthermore, different robots will be used to validate the generic aspect of the bio-inspired controller: the Franka/Panda robot arm at the LORIA [Menges et al, 2021], the Baxter bi-arm robot at Kyutech [Yamasaki et al. 2023]. Human movements and EEG will be recorded to measure the synchronization by the phase-lock value (PLV) index [Jouaiti et al. 2018] and the emergence of motor coordination between the human and the robot.
Handshaking (HS)
HS is a rhythmic interaction with physical contact and mechanical coupling by grasping hands. It is a central, common, and non-verbal communication act combining discrete and rhythmic movements, which opens/closes, regulates, and sets the stage for the communication between two individuals [Schiffrin, 1974, Giannopoulos et al. 2011]. It involves a fine and complex coordination of hand, body and also gazes [Walker et al, 2010]. In this project, we will focus only on the movement coordination between humans and robots (without gaze and speech interaction) and analyze the emotional state (affect) of the human and its engagement depending on the robot’s rigidity/smoothness and its ability to synchronize with the human gesture. The robot will be controlled by the same CPG networks as RJ. Human movements, heart rate, and EEG will be recorded to measure the synchronization, the emergence of motor coordination, the human affect and engagement when interacting with the robot.
HexGame (HG)
Joint attention (JA) is a fundamental skill for human social cognition, allowing individuals to focus on things together with others. Through this capacity, humans can share experiences about the world, coordinate thoughts and behaviors, and successfully cooperate with others. Reduced JA is associated with deficits in social cognition, which is observed in psychological conditions such as autism spectrum disorders (ASD).
By continuing a previous study where a model is proposed to track JA in human-robot interaction1, the objective of this project is to develop an experimental prototype for JA skill training, consisting in playing Hex-game against the robot iCub. Through the prototype developed, the following hypothesis are going to be evaluated:
a) Shorter decision-making time is obtained by engaging in JA with iCub from gaze cueing.
b) The game situation and embodiment can help training JA skills in ASD persons.