The inner speech is thoroughly studied in humans, and it represents an interdisciplinary research issue involving psychology, neuroscience, and pedagogy. A few papers only, mostly theoretical, analyze the role of inner speech in robots. The present study investigates the potential of the robot's inner speech while cooperating with human partners. A cognitive architecture is designed and integrated with standard robot routines into a complex framework. Two threads of interaction are discussed by setting the robot operations with and without inner speech. Thanks to the robotic self-dialog, the partner can easily trace the robot's processes. Moreover, the robot can better solve conflicts leading to successful goal achievements. The results show that functional and transparency requirements, according to the international standards ISO/TS:2016 and COMEST/Unesco for collaborative robots, are better met when inner speech accompanies human-robot interaction. The inner speech could be applied in many robotics contexts, such as learning, regulation, and attention.

Inner Speech is an essential but also elusive human psychological process that refers to an everyday covert internal conversation with oneself. We argued that programming a robot with an overt self-talk system that simulates human inner speech could enhance both human trust and users’ perception of robot’s anthropomorphism, animacy, likeability, intelligence and safety. For this reason, we planned a pre-test/post-test control group design. Participants were divided in two different groups, one experimental group and one control group. Participants in the experimental group interacted with the robot Pepper equipped with an over inner speech system whereas participants in the control group interacted with the robot that produces only outer speech. Before and after the interaction, both groups of participants were requested to complete some questionnaires about inner speech and trust. Results showed differences between participants’ pretest and post-test assessment responses, suggesting that the robot’s inner speech influences in participants of experimental group the perceptions of animacy and intelligence in robot. Implications for these results are discussed.

As Artificial Intelligence and Robotics evolve, the ethical implications of autonomous systems are becoming increasingly paramount. This article explores the role of a robot’s inner speech in enhancing human phronesis - the capacity for making ethical and contextually appropriate decisions. Phronesis is a complex human trait based on experience, personality, and values, and is crucial for decisions affecting others’ well-being. While artificial systems (i.e., robots) may not fully replicate human practical wisdom, they can support humans by stimulating reflection and helping in situations requiring careful consideration. This support can be more impactful if the robot has the ability for inner speech. The article examines whether a robot engaging in internal dialogue can help develop phronesis in people interacting with it, especially in contexts where caring of people with special needs is crucial. A group of participants was involved in a virtual game, where they had to care for a suffering person alongside a robot. Their decisions, in terms of task resolution and ethical considerations, allowed to assess if the robot’s inner speech increased participants’ awareness of the suffering person’s needs. Preliminary results are promising, highlighting the potential of thinking robots to assist in making ethical decisions.

The robot’s inner speech has recently been under investigation. Inner speech is a form of internalized, self-directed dialogue: talking to the self. Initial intriguing findings pertained to the robot’s transparency, robustness, and trustworthiness, and emerged from basic collaborative tasks in Human-Robot Interaction (HRI), such as set up a lunch table. This pilot study extends the investigation to a high-stakes medical scenario where errors can have significant consequences: to prepare a virtual surgical table for a vascular intervention. The study involved both expert and non-expert participants who collaborated with the robot to correctly place and organize surgical instruments. Data were collected through a post-session questionnaire combining closed-ended Likert-scale items and open-ended questions. Interaction dynamics and learning outcomes were assessed through key variables, including transparency, trust, reassurance, and perceived learning, allowing for both quantitative and qualitative analysis. Results from this pilot indicate that the robot’s inner speech enhances participants’ understanding of task instructions, provides reassurance during collaboration, and improves perceived learning quality, evaluated in terms of the Quality of Learning (QoL) framework, with effects varying according to prior expertise. These findings highlight the potential of inner speech to facilitate transparent, trustworthy, and effective training in risk-sensitive domains, offering a foundation for further exploration of robotic tutoring systems in healthcare.