The introduction of social robots in preschool settings has become a common research strategy for addressing educational challenges. Although teachers and educators play a central role in classroom dynamics, they are often underrepresented in studies on educational robots. Often, robots are presented as “black-boxes”, with little attention paid to providing teachers with dedicated training. This study describes the design and implementation of the Teacher Professional Development Research (TPDR) as a structured method for integrating social robots into early education, supporting teachers and educators. TPDR is an established educational practice that addresses pedagogical issues by engaging teachers as actors in the research process. Our project involved deploying a robot in four preschools and one nursery with a multicultural setting, primarily to foster intercultural integration. Both quantitative and qualitative data were collected to evaluate the impact of this approach on teachers' attitudes and willingness to adopt the robot. Findings indicate that the teachers gained a greater awareness of the robot’s social presence and a clearer understanding of its educational potential. There was also an overall positive shift in their intercultural sensitivity.

The objective of this work is to develop and evaluate computational cognitive models of Theory of Mind (ToM) and Machiavellian behavior embedded in a humanoid robot. Machiavellianism, together with psychopathy and narcissism, is part of the Dark Triad (DT), three constructs that correspond to socially aversive yet not necessarily pathological personalities. The motivations of the present work are both theoretical and application-oriented. In the long term, we aim to: (i) Provide researchers with new insights into the Machiavellian as well as other DT constructs through simulated and robotic setups; (ii) Provide a tool to train psychologists to deal with social and antisocial behavior in a controlled setup; (iii) Help people become aware of the behavioral mechanisms that they may expect from people with DT traits in social and affective relationships; (iv) Assist robotic engineers in developing better robots by identifying behaviors that should be avoided. To this end, we explored a computational model of ToM in the popular Planning Domain Definition Language (PDDL), and defined a domain with the necessary elements to induce Machiavellian behavior during planning and execution. Subsequently, we implemented our computational model in a software architecture controlling the behavior of a humanoid robot and recorded videos of the robot interacting with two actors. Finally, we conducted experiments with 300 participants divided into 6 conditions to verify whether the implemented framework is versatile enough to generate behaviors that participants would rate as either more Machiavellian or less Machiavellian based on their observations of the recorded videos.

Co-design is widely used in educational contexts to involve stakeholders and make them active participants in the learning process. This study presents the co-design process conducted with teachers, educators, and families before introducing a social robot in four Italian preschools and a nursery. The robot is expected to promote intercultural awareness in a highly culturally diverse educational environment. We consider the co-design process an essential step, as teachers and educators, by knowing the social rules and pedagogical concepts of each specific educational context, can tailor the unique characteristics of the robot (being embodied and equipped with social behaviors) to effectively benefit their pedagogical reality. In addition, stakeholders, through co-design, can incorporate cultural awareness of children and their families into the robot design. The results obtained after the co-design process highlight that the co-creation of robotic applications and the robot’s imagery before its actual introduction into activities is fundamental for developing a framework tailored to a specific educational context, for reformulating the project’s prerogatives in such a way that it becomes part of the educational reality, and for giving teachers the opportunity to familiarize themselves with the robot, understand its capabilities, and exploit them according to their educational context.

This work explores a novel approach to empowering robots with visual perception capabilities using textual descriptions. Our approach involves the integration of GPT-4 with dense captioning, enabling robots to perceive and interpret the visual world through detailed text-based descriptions. To assess both user experience and the technical feasibility of this approach, experiments were conducted with human participants interacting with a Pepper robot equipped with visual capabilities. The results affirm the viability of the proposed approach, allowing to perform vision-based conversations effectively, despite processing time limitations.

The synergic orchestration of the cognitive and psychological dimensions characterizes human intelligence. Accordingly, carefully designing this mechanism in artificial intelligence can be a successful strategy to increase human likeness in a robot, enhancing mutual understanding and building a more natural and intuitive interaction. For this purpose, the main contribution of this work is a psychological and cognitive architecture tailored for HRI based on the interplay between robotic personality and memory-based cognitive processes. Indeed, the artificial personality manifests itself not only in various aspects of the behavior but also within the action selection process, which is closely intertwined with personality-dependent hedonic experiences linked to memories. Within this paper, we propose a task- and platform-independent framework, evaluated in a multiparty collaborative scenario. Obtained results show that a robot connected to our proposed framework is perceived as a cognitive agent capable of manifesting perceivable and distinguishable personality traits.

This work illustrates the testing of the Spot Robot performed at the training camp of Civil Protection and ANPAS (National Association of Public Assistance) in Foligno. The camp simulates the aftermath of an earthquake with different types of collapsed buildings. We teleoperated the quadruped Spot robot in different areas of the camp where Spot needs to address different challenges. The focus of the testing was not on the objective performance of the robot but on how the robot was subjectively perceived by rescuers of ANPAS and Civil Protection. Initially, we formulated and tested two hypotheses to check if locomotion in some areas is perceived better than in other areas and if there are perceivable differences when the robot is using different types of locomotion gaits. Then, we conducted unstructured interviews with participants who observed the robot in action to describe their rescue procedures and give us suggestions and opinions on what operations they expect the robot might perform.

This study introduces a system implemented on a legged robot, designed to generate a multi-layered map that incorporates semantic information, specifically tailored for Search & Rescue robotics. The article discusses the development of a Machine Learning model based on visual data for recognizing people and environmental features, and its integration into a mapping and navigation architecture. The system was tested in two different locations using the Spot robot by Boston Dynamics, equipped with an external ZED2 depth camera. Tests are described in detail and results analyzed.

Co-speech gesture generation on artificial agents has gained attention recently, mainly when it is based on data-driven models. However, end-to-end methods often fail to generate co-speech gestures related to semantics with specific forms, i.e., Symbolic and Deictic gestures. In this work, we identify which words in a sentence are contextually related to Symbolic and Deictic gestures. Firstly, we appropriately chose 12 gestures recognized by people from the Italian culture, which different humanoid robots can reproduce. Then, we implemented two rule-based algorithms to label sentences with Symbolic and Deictic gestures. The rules depend on the semantic similarity scores computed with the RoBerta model between sentences that heuristically represent gestures and sub-sentences inside an objective sentence that artificial agents have to pronounce. We also implemented a baseline algorithm that assigns gestures without computing similarity scores. Finally, to validate the results, we asked 30 persons to label a set of sentences with Deictic and Symbolic gestures through a Graphical User Interface (GUI), and we compared the labels with the ones produced by our algorithms. For this scope, we computed Average Precision (AP) and Intersection Over Union (IOU) scores, and we evaluated the Average Computational Time (ACT). Our results show that semantic similarity scores are useful for finding Symbolic and Deictic gestures in utterances.

This paper presents a system for diversity-aware autonomous conversation leveraging the capabilities of large language models (LLMs). The system adapts to diverse populations and individuals, considering factors like background, personality, age, gender, and culture. The conversation flow is guided by the structure of the system’s pre-established knowledge base, while LLMs are tasked with various functions, including generating diversity-aware sentences. Achieving diversity-awareness involves providing carefully crafted prompts to the models, incorporating comprehensive information about users, conversation history, contextual details, and specific guidelines. To assess the system’s performance, we conducted both controlled and real-world experiments, measuring a wide range of performance indicators.

The goal-directed behavior observed in humans arises from the intricate interplay of various processes, including personality dynamics, emotional responses to others, memory encoding, the anticipation of future actions, and associated hedonic experiences. Integrating these multiple processes characteristic of human intelligence into a robotic framework aims to enhance the human-likeness of artificial agents and facilitate more natural and intuitive interactions with humans.For this purpose, in this paper, we propose a comprehensive psychological and cognitive architecture where, personality, as it happens for humans, not only influences the execution of actions but also shapes internal reactions to human emotions and guides anticipatory decision-making processes tailored to the agent’s traits. We demonstrate the framework’s effectiveness in generating perceivable synthetic personalities through an experiment involving participants in a dyadic conversation scenario with a digital human, where the digital human’s behavior is driven by its assigned personality. The results show that participants accurately perceive the artificial personality displayed by the digital human. We also demonstrate the potential of our robotic framework to bridge the gap between cognitive and psychological agents, as the findings highlight its ability to create a cognitively and emotionally intelligent digital human.

In the future, eventually, robots will become extremely widespread also in urban environments, and perhaps, us humans will need to learn how to interact and live with them. Social navigation accounts for the problem of having a safe and efficient navigation among objects and pedestrians, which can be considered as sentient road users and, for this reason, more special considerations need be taken into account when dealing with them. The goal of any social navigation software stack is to make the robotic agent behave as similarly as possible to a pedestrian, which is used to abide to many social rules that has learnt throughout all of their life. In this way, humans will not need to learn new "robotic" rules for navigating an environment: they would only need to apply the same rules that also robots will follow. Many social navigation approaches rely on sociological-psychological studies in which the pedestrian motion has been modeled in deep details. In this work a novel approach is presented, leveraging the predictivity of Model Predictive Control and the reactivity of Social Force Model, which will model the pedestrian motion.

Educational robotics is increasingly merging into school curricula. Understanding the subjective perceptions of robots, particularly among children, requires nuanced approaches. In this paper, we employed a qualitative coding analysis method to explore how children of different ages conceptualise robots through drawings, and how prior experiences with robotics influence their perceptions. Our findings reveal that the perception of robotics is influenced by cognitive development stages, which is in turn affected by age, and by educational robotics. The latter plays a significant role in shaping perceptions of robots, fostering positive attitudes and aiding cognitive development, particularly in first-grade students. Our insights can inform both teachers to better tailor their educational robotics activities for different age groups and robot designers themselves. For instance, our findings highlight the importance of emotional expression and the preference for humanoid robots among primary school children.

This work describes an immersive control system for a quadruped robot, designed to track the head movements of the operator wearing a virtual reality eye-wear, while also utilizing joystick commands for locomotion control. The article details the implemented closed-loop velocity control approach and the locomotion task specifications. The proposed method has been implemented on a Spot robot from Boston Dynamics, with Meta Quest 2 virtual reality system. Evaluation of the approach involved a user study, where participants engaged in immersive control of the quadruped robot within an indoor experimental environment and provided feedback through standardized questionnaires. Pairwise comparison of the resulting data revealed significant advantages for the proposed immersive control system over a standard remote controller, with enhanced performance observed in the second trial of using the control system. However, participants lacking experience with virtual reality systems reported increased distress symptoms following the experiment.Code: https://www.github.com/aliy98/zed-oculus-spot

The purpose of this research is to study the impact of robot participation in group conversations and assess the effectiveness of different addressing policies. The study involved a total of 300 participants, who were divided into groups of four and engaged in a dialogue with a humanoid robot. The robot acted as a moderator, using information obtained during the conversation to determine which speaker to address. The study found that the policy used by the robot significantly impacted the conversation dynamics. Specifically, the robot provided more balanced attention to each participant and reduced the number of subgroups.

In this work, we propose a strategy to implement the first two steps of the DRABC paradigm (Danger, Response, Airway, Breathing, Circulation) used by rescuers in Search and Rescue (SAR) with the use of a mobile quadruped robot. The robot is programmed to autonomously explore and create a map of the environment with the main objective of identifying areas of danger and reporting them to rescuers (first step of DRABC). While completing this first goal the robot must also identify people still inside the building, mark their position but also evaluate the health state of the person and in particular the response (second step of DRABC). Specifically, we propose new strategies for SAR considering that autonomous behaviour is particularly relevant before the human rescuers arrive: therefore, the policy adopted should privilege covering a broader area in the available time, rather than exploring a smaller area in depth. Strategies have been tested with the Spot robot from Boston Dynamics concerning both exploration and health assessment. The software developed and the tests to validate it are thoroughly described and explained.

The objective of this work is to develop computational cognitive models embedded in a humanoid robot. We focus on Dark Triad constructs and the so-called “Nice and Nasty” Theory of Mind that have never been investigated through a robotic approach. To this end, DT and ToM conceptual models in psychology have been taken as a reference for developing a framework based on the popular PDDL planning language. Next, a cognitive architecture has been implemented on a humanoid robot, with the final objective of making adverse personalities emerge. The motivations of the present work are both theoretical and practical. On the one side, we aim to provide researchers with new insights into DT constructs through simulated and robotic setups. On the other side, we aim to provide a tool to train psychologists to deal with social and antisocial behaviour in a controlled setup. The article includes all the details about the model and the experiments performed.

This article explores the acceptance of a humanoid robot designed to engage in conversations with clinicians and individuals with spinal cord injuries. Building upon prior research, we introduce the concept of “diversity-aware” robots, which possess the capability to interact with people while adapting to their culture, age, gender, preferences, and physical and mental conditions. These robots are connected to a cloud system specifically designed to consider these factors, enabling them to adapt to the context and individuals they interact with. Our experiments involved the NAO robot interacting with both clinicians and individuals with spinal cord injuries in a hospital environment. Subsequent to the interaction, participants completed a questionnaire and underwent an interview. The collected data were analyzed to assess the system’s acceptability and its persistence beyond the initial novelty effect. Furthermore, we investigated whether clinicians exhibited a lower predisposition towards the system and expressed greater concerns than end-users about using the robot, which could potentially hinder the adoption of the system.

Robotic personalities broaden the social dimension of an agent creating feelings of comfort in humans. In this work, we propose a taxonomy model to generate synthetic personalities based on the Big Five model. In particular, this paper describes a generalized framework for artificial personalities whose core is a Bidirectional Encoder Representations from Transformers (BERT) model capable of associating behaviors tailored to each personality trait. The generator is fully integrated within a modular software architecture capable of performing social interaction tasks, being at the same time task-and platform-independent. The proposed framework has been tested in a pilot experiment where human subjects were asked to interact with a humanoid robot displaying different personality traits. Results obtained by the statistical analysis of validated questionnaires show interesting insights about the capability of the framework of generating personalities that are clearly perceived by users, and whose personality dimensions are strongly distinguishable.

This article proposes a new framework to model a scenario in which First Responders, citizens, smart devices, or robots explore the environment in an emergency situation, i.e., after an earthquake, assessing damages and searching for people needing assistance. While moving, the agents observe events and exchange the information collected with other agents encountered: to this end, they use messaging systems purposely adapted to use point-to-point network connections to allow local data exchange between agents even when global network connections are not available. As is common in Delay Tolerant Networks, exchanged messages are locally stored: when a global network is available, the agents can upload all the information collected by themselves and other agents they encountered to a Control Room or a database in the Cloud. Differently from traditional DTN algorithms such as Epidemic and Spray&Wait, we propose a solution that keeps track of agents that shared information along the path and assess the quality of the information collected by multiple agents through a reputation-based mechanism that is safer than majority voting. A simulator compatible with OpenStreetMap is presented, as well as simulated experiments in two Italian towns to validate the feasibility of the approach.

Emotions have always played a crucial role in human evolution, improving not only social contact but also their ability to adapt and react to a changing environment. In the field of social robotics, providing robots with the ability to recognize human emotions through the interpretation of non-verbal signals may represent the key to more effective and engaging interaction. However, the problem of emotion recognition has usually been addressed in limited and static scenarios, by classifying emotions using sensory data such as facial expressions, body postures, and voice. This work proposes a novel emotion recognition framework, based on the appraisal theory of emotion. According to the theory, the expected person’s appraisal of a given situation depending on their needs and goals (henceforth referred to as "appraisal information") is combined with sensory data. A pilot experiment was designed and conducted: participants were involved in spontaneous verbal interaction with the humanoid robot Pepper, programmed to elicit different emotions in various moments. Then, a Random Forest classifier was trained to classify positive and negative emotions using: (i) sensor data only; (ii) sensor data supplemented by appraisal information. Preliminary results confirm a performance improvement in emotion classification when appraisal information is considered.

Achieving natural and engaging verbal interactions is one of the main challenges faced by Social Robotics. In this context, physical embodiment may be one of the most critical factors: indeed, previous work indicates that physical robots elicit more favorable social responses than virtual agents. However, the effects of physical embodiment have been analysed only in some specific and limited scenarios, where verbal interaction was reduced to basic commands.The current work aims at investigating the effect of robots' physical embodiment in a pure conversation task, by considering some relevant aspects of social interaction, such as usability, speech interface quality, user satisfaction and engagement. To this aim, a pilot experiment where participants were required to chitchat with a robot and a smartphone app, both connected to the same conversation framework, has been carried out. Preliminary results are presented and discussed, and they offer interesting insights about the positive effects of physical embodiment on some of the analysed aspects.

Unmanned Aerial Vehicles (UAVs) popularity is increased in recent years, and the domain of application of this new technology is continuously expanding. However, although UAVs may be extremely useful in monitoring contexts, the operational aspects of drone patrolling services have not yet been extensively studied. Specifically, patrolling and inspecting with UAVs different targets distributed over a large area is still an open problem, due to battery constraints and other practical limitations. In this work, we propose a deterministic algorithm for patrolling large areas in a pre- or post-critical event scenario. The autonomy range of UAVs is extended with the concept of Social Drone Sharing: citizens may offer their availability to take care of the UAV if it lands in their private area, being thus strictly involved in the monitoring process. The proposed approach aims at finding optimal routes in this context, minimizing the patrolling time and respecting the battery constraints. Simulation experiments have been conducted, giving some insights about the performance of the proposed method.

Personality is one of the most important factors in human interactions, which retains its importance in human- robot interactions with social robots. This work focusses on the varied linguistic strategies which characterize the personality traits of extraversion and introversion, analysing the main features that differentiate both personalities and eventually proposing a verbal communication model for the extraverted and introverted personality of a conversational social robot. The model classifies and converts phrases, with the result of building different communication styles. A pilot study, involving human subjects and the humanoid robot Pepper, programmed to mimic both extraverted and introverted personality types, has been conducted, with the twofold aim of assessing if differences between the two personalities of the robot can be perceived, and analyzing the effects of different personality traits on verbal interaction with human subjects. Preliminary results seem to confirm the law of attraction in human-human interaction for the extraverted personality.

The article focusses on the work performed in preparation for an experimental trial aimed at evaluating the impact of a culturally competent robot for care home assistance. Indeed, it has been estabilished that the user's cultural identity plays an important role during the interaction with a robotic system and cultural competence may be one of the key elements for increasing capabilities of socially assistive robots. Specifically, the paper describes part of the work carried out for the definition and implementation of two different robotic systems for the care of older adults: a culturally competent robot, that shows its awareness of the user's cultural identity, and a reference robot, non culturally competent, but with the same functionalities of the former. The design of both robots is here described in detail, together with the key elements that make a socially assistive robot culturally competent, which should be absent in the non-culturally competent counterpart. Examples of the experimental phase of the CARESSES project, with a fictional user are reported, giving a hint of the validness of the proposed approach.

The way in which humans behave, speak and interact is deeply influenced by their culture. For example, greeting is done differently in France, in Sweden or in Japan; and the average interpersonal distance changes from one cultural group to the other. In order to successfully coexist with humans, robots should also adapt their behavior to the culture, customs and manners of the persons they interact with. In this paper, we deal with an important ingredient of cultural adaptation: how to generate robot plans that respect given cultural preferences, and how to execute them in a way that is sensitive to those preferences. We present initial results in this direction in the context of the CARESSES project, a joint EU-Japan effort to build culturally competent assistive robots.

The functionalities and behaviours of socially assistive robots for the care of older people are usually defined by the robot's designers with limited room for runtime adaptation to meet the preferences, expectations and needs of the assisted person. However, adaptation plays a crucial role for the robot's acceptability and ultimately for its effectiveness. Culture, which deeply influences a person's preferences and habits, can be viewed as an invaluable “enabling technology” to achieve such level of adaptation. This paper discusses how guidelines describing culturally competent assistive behaviours can be encoded in a robot to effectively tune its actions, gestures and words. The proposed system is implemented on a Pepper robot and tested with an Indian persona, whose habits and preferences the robot discovers and adapts to at runtime.

Wheeled Robots (WRs) are widely used in many different contexts, and usually they are required to operate in partial or total autonomy. In particular, in a wide range of situations, having the capability of following a predetermined path and avoiding unexpected obstacles can be extremely relevant. On these basis, this paper analyzes an integrated approach for path following and obstacle avoidance applied to unicycle-type robots. The approach is based on the definition of the path to be followed as a curve f(x, y) in space, while obstacles are modeled as Gaussian functions that modify the original function, generating a resulting safe path. The attractiveness of this methodology which makes it look very simple, is that it neither requires the computation of a projection of the robot position on the path, nor does it need to consider a moving virtual target to be tracked. The performances of the proposed approach are analyzed by means of a series of experiments performed in dynamic environments with unicycle-type robots.

The article proposes a solution allowing a multicopter to generate and follow a path while taking into account the obstacles in the environment. Specifically, we introduce a method for path definition that describes a curve as the intersection of two surfaces. Then, the article proposes a computationally efficient strategy allowing to modify either surface, and hence the resulting path, to take into account the presence of obstacles perceived in real-time. The algorithm has been implemented and embedded in a software package to control the flight of a fully autonomous AscTec Firefly hexacopter with two cameras and onboard processing capabilities.

A common way to organize a high number of robots, both when moving autonomously and when controlled by a human operator, is to let them move in formation. This is a principle that takes inspiration from the nature, that maximizes the possibility of monitoring the environment and therefore of anticipating risks and finding targets. In robotics, alongside these reasons, the organization of a robot team in a formation allows a human operator to deal with a high number of agents in a simpler way, moving the swarm as a single entity. In this context, the typology of visual feedback is fundamental for a correct situational awareness, but in common practice having an optimal camera configuration is not always possible. Usually human operators use cameras on board the multirotors, with an egocentric point of view, while it is known that in mobile robotics overall awareness and pattern recognition are optimized by exocentric views. In this article we present an analysis of the performance achieved by human operators controlling a swarm of UAVs in formation, accomplishing different tasks and using different point of views. The control architecture is implemented in a ROS framework and interfaced with a 3D simulation environment. Experimental tests show a degradation of performance while using egocentric cameras with respect of an exocentric point of view, although cameras on board the robots allow to satisfactorily accomplish simple tasks.