As home service robots become increasingly integrated into domestic environments, trust in Human–Robot Interaction (HRI) emerges as a critical factor influencing their acceptance and effectiveness. This article presents survey review aiming to provide an understanding and insights into trust in HRI, particularly in home service robots. By analyzing existing studies, we explore the definition of trust and dimensions in different disciplinary perspectives, influencing factors, and assessment methodologies of trust in HRI. We also explore the dynamic nature of trust, highlighting the roles of robot reliability, transparency, predictability, and social interaction in shaping user perceptions. Furthermore, we present existing trust measurement approaches, including self-report questionnaires, behavioral and physiological metrics, and multimodal assessments, while identifying gaps in standardization and real-time evaluation. Ethical considerations, cultural influences, and the long-term evolution of trust in home service robots are also discussed. By synthesizing insights from interdisciplinary research, this article aims to provide insights that will benefit both the academic and research communities, as well as practical applications, and to inform future studies and guide the development of trustworthy, adaptive and user-centered home service robots that seamlessly integrate into daily life.

International audience

Predictive maintenance is a critical approach in various industries to enhance operational efficiency and minimize downtime by forecasting equipment failures. Publicly available datasets have been widely used to evaluate predictive maintenance algorithms, but they do not include complexities such as feet context, missing data, and irregular data acquisition. This paper presents a comparative analysis of public datasets and introduces a new one AI4I-PMDI to overcome the shortcomings of available datasets, which fail to accurately represent realistic maintenance data. AI4I-PMDI is derived from the AI4I 2020 predictive maintenance dataset, which is enhanced with irregularities that are common in real maintenance data. Moreover, machine learning algorithms were applied to both the original public dataset and the modified dataset for binary and multi-class Classification tasks. The result revealed that the modifications applied significantly impact the performances. This emphasizes the importance of using datasets adapted to the constraints encountered in complex industrial settings.

International audience

International audience

The utilization of virtual reality (VR) technology has shown promise in various therapeutic applications, particularly in exposure therapy for reducing fear of certain situations objects or activities, e.g. fear of height, or negative evaluation of others in social situations. VR has been shown to yield positive outcomes in follow-up studies, and provides a safe and ecological therapeutic environment for therapists and their patients. This paper presents a collaborative effort to develop a VR speech therapy system which simulates a virtual audience for users to practice their public speaking skills. We describe a novel web-based graphical user interface that enables therapists to manage the therapy session using a simple timeline. Lastly, we present the results from a qualitative study with therapists and teachers with functional dysphonia, which highlight the potential of such an application to support and augment the therapists' work and the remaining challenges regarding the design of natural interactions, agent behaviours and scenario customisation for patients.

The use of deep learning in semantic segmentation of point clouds enables a drastic improvement of segmentation precision. However, available datasets are restrained to a few applications with limited applicability to other fields. Using synthetic and real data can alleviate the burden of creating a dedicated dataset at the cost of domain-shift that is mostly addressed during training, while treating the problem directly on the data has been less explored. Towards this goal, two methods to alleviate domain shift are proposed, firstly by enhanced generation and sampling of synthetic data and secondly by leveraging color information of unlabeled point clouds to color synthetic, uncoloured data. Obtained results confirm their usefulness in improving semantic segmentation result (+3.43 into mIoU for a network trained on S3DIS zone 1). More importantly, the devised coloring method shows the ability of a point-based network to link color information with recurrent geometric features. Finally, the presented methods are able to bridge the domain-shift gap even in cases where inclusion of raw synthetic data during training impedes learning.

Point cloud acquisition techniques are an essential tool for the digitization of industrial plants, yet the bulk of a designer's work remains manual. A first step to automatize drawing generation is to extract the semantics of the point cloud. Towards this goal, we investigate the use of deep learning to semantically segment oil and gas industrial scenes. We focus on domain characteristics such as high variation of object size, increased concavity and lack of annotated data, which hampers the use of conventional approaches. To address these issues, we advocate the use of synthetic data, adaptive downsampling and context sharing.

This paper presents the first prototype of a virtual audience system (VAS) specifically designed as a training tool for conference talks. This system has been tailored for university seminars dedicated to the preparation and delivery of scientific talks. We describe the required features which have been identified during the development process. We also summarize the preliminary feedback received from lecturers and students during the first deployment of the system in seminars for bachelor and doctoral students. Finally, we discuss future work and research directions. We believe our system architecture and features are providing interesting insights on the development and integration of VR-based educational tools into university curriculum.

Virtual training environments experienced with different immersive technologies can accommodate users' preferences, proficiency, and platform availability. Whereas research comparing the effects of immersive technologies can provide important insights about their impact on users' experience (e.g. engagement, transfer of learning), current studies do not address how to design the user interface (UI) to ensure sound comparisons across platforms. For effective comparisons, however, the UI designs must be adapted for the platform used to provide comparable usability. In this article we describe our UI design methodology for the development of an effective and usable virtual classroom training simulator built for three technologies: (1) desktop; (2) Head-Mounted Display (HMD); and (3) Cave Automatic Virtual Environment (CAVE). Usability and other user experience factors were evaluated for each platform with concurrent think-aloud protocol and semi-structured interviews indicating that all three UIs were easy to use and to learn. We discuss insights for future development of cross-platform VTEs.

To be capable of life-long learning in a real-life environment, robots have to tackle multiple challenges. Being able to relate physical properties they may observe in their environment to possible interactions they may have is one of them. This skill, named affordance learning, is strongly related to embodiment and is mastered through each person's development: each individual learns affordances differently through their own interactions with their surroundings. Current methods for affordance learning usually use either fixed actions to learn these affordances or focus on static setups involving a robotic arm to be operated. In this article, we propose an algorithm using intrinsic motivation to guide the learning of affordances for a mobile robot. This algorithm is capable to autonomously discover, learn and adapt interrelated affordances without pre-programmed actions. Once learned, these affordances may be used by the algorithm to plan sequences of actions in order to perform tasks of various difficulties. We then present one experiment and analyse our system before comparing it with other approaches from reinforcement learning and affordance learning.

Elementary school classrooms are emotionally stressful environments, for both students and teachers. Successful teachers use strategies that regulate students' emotions and behaviors while also controlling their own emotions (stress, nervousness). To prepare teachers for the challenges of teaching, teacher training should include emotional and behavioral management strategies. Virtual Training Environments (VTEs) are effective at providing experiences and increasing learning in many domains. Creating VTEs for teachers can improve student learning and teacher retention. We introduce our current research aimed at integrating emotionally-intelligent virtual students within a 3D classroom training system. In our simulation, virtual students' emotional states will be determined from an appraisal process of actions taken by the teacher trainee in the virtual classroom. Virtual students will then display the appropriate non-verbal behaviors and react to the teacher accordingly. We present the first steps required to implement our proposed architecture which are based on appraisal theory of emotions and emotion regulation theory.

Simulation-based training systems have proven effective in a variety of domains, both for facilitating the learning of skills as well for applying this knowledge to real life. Although difficulties managing students' disruptive behavior in classrooms has been identified as one of the main causes of teachers' turnover, only a handful of virtual training environments have focused on providing training to teachers, and still no clear methodologies exist for their design, their implementation, nor their evaluation.

In this paper, we describe the implementation and performance of a Virtual Audience perception model for Virtual Reality (VR). The model is a VR adaptation of an existing desktop model. The system allows a user in VR to easily build and experience a wide variety of atmospheres with small or large groups of virtual agents.The paper describes results of early evaluations for this model in VR. Our first scalability benchmark results demonstrated the ability to simultaneously handle one hundred virtual agents without significantly affecting there commended frame rate for VR applications.This research is conducted in the context of a classroom simulation software for teachers’ training.

In this paper, we present a 3D virtual environment for inducing body ownership illusion. The key idea is to take advantage of virtual reality potential regarding body size perception. The application, called REVAM, links two main aspects of body size perception: the first one focuses on its modification and the second one concerns its assessment. Based on some previous evidence that it is possible to modify body-size perception through an illusion of ownership over a virtual body, the application proposes to couple a tactile stimulation when viewing an avatar from a third person perspective (a condition known to produce this kind of illusion). In addition, the application offers the possibility to choose between avatars of different builds, and to perform morphing to reduce the avatars body. Moreover, the application allows to implicitly measure how people perceive their body size from an affordance estimation task in which people have to appreciate if they can pass through doors of different sizes without twisting their shoulders. To test the application we carried out an experiment on 16 female participants who performed the affordance estimation task five times: the first time before being exposed to their chosen avatar to get a baseline measure, and the four other times after being exposed to their avatar in different situations. These different situations are defined by the crossing of two experimental factors: morphing (presence or absence) and simultaneous visuotactile stimulation (presence or absence). A two-way repeated-measures Anova showed a main effect of the morphing: mean door width through which the participants estimate they can fit was significantly reduced (p <;:05) when morphing was present. However, this effect did not interact with the presence of a simultaneous tactile stimulation. This indicates that exposing people to a virtual body reduced in size, as proposed in the present application, could be an effective way to modify body size perception, at least temporarily. The final goal would be to help patients with body image disorders, such as anorexia nervosa.

This paper introduces the design of autonomous virtual player based on imitation learning using human behavior observations. The ORION model provides both data mining techniques allowing the extraction of knowledge and behavior models allowing the control of the autonomous behaviors. ORION is also an operational tool allowing the representation, transformation, visualization and prediction of data. We illustrate the use of our model by detailing the implementation of a virtual player for the video game Unreal Tournament 3. Thanks to ORION, data from low level behaviors were collected through three scenarios performed by human players: movement, long range aiming and close combat. Behaviors can then be learned from the obtained data-sets after transformations and application of data mining techniques. ORION allows us to build a complete behavior using an extension of a Behavior Tree integrating ad hoc features in order to manage aspects of behavior that we have not been able to learn automatically.

For video game designers, being able to provide both interesting and human-like opponents is a definite benefit to the game's entertainment value. The development of such believable virtual players also known as Non-Player Characters or bots remains a challenge which has kept the research community busy for many years. However, evaluation methods vary widely which can make systems difficult to compare. The BotPrize competition has provided some highly regarded assessment methods for comparing bots' believability in a first person shooter game. It involves humans judging virtual agents competing for the most believable bot title. In this paper, we describe a system allowing us to partly automate such a competition, a novel evaluation protocol based on an early version of the BotPrize, and an analysis of the data we collected regarding human judges during a national event. We observed that the best judges were those who play video games the most often, especially games involving combat, and are used to playing against virtual players, strangers and physically present players. This result is a starting point for the design of a new generic and rigorous protocol for the evaluation of bots' believability in first person shooter games.

Virtual reality is a frequently used tool in vocational training. Nevertheless, its efficiency has not been systematically tested. The main goal of this study is to assess the effectiveness of a virtual environment (VE) for learning a complex procedure in the biomedical domain. Two experiments were performed. The first one assessed internal validity of the VE, which is the effectiveness of using a VE in the process of learning a new procedure. The second one tested external validity of the VE, which is the participants’ ability to reproduce the acquired skills in a real context. We find that internal and external validity must be evaluated before using a virtual environment in long-term procedure learning. The results of such evaluations are a basis for future experiments aiming to optimize learning conditions in a VE and transferring the acquired skills in a real context.

Abstract This paper is dedicated to the study of existing approaches that explicitly use mental simulation. Current implementations of the mental simulation paradigm, taken together, computationally address many aspects suggested by cognitive science research. Agents are able to find solutions to nontrivial scenarios in virtual or physical environments. Existing systems also learn new behavior by imitation of others similar to them and model the behavior of different others with the help of specialized models, culminating with the collaboration between agents and humans. Approaches that use self models are able to mentally simulate interaction and to learn about their own physical properties. Multiple mental simulations are used to find solutions to tasks, for truth maintenance, and contradiction detection. However, individual approaches do not cover all of the contexts of mental simulation and most rely on techniques which are only suitable for subsets of obtainable functionality. This review spans through four perspectives on the functionality of state‐of‐the‐art artificial intelligence applications, while linking them to cognitive science research results. Finally, an overview identifies the main gaps in existing literature on computational mental simulation and provides our suggestions for future development.

We investigate the decision-making and behavior of robotic biped goalkeepers, applied to the RoboCup 3D Soccer Simulation League. We introduce two approaches to the goalkeeper’s behavior: first a heuristics-based approach that uses linear regression and Kalman filters for improved perception, and another based on mental models which uses nonlinear regression for ball trajectory filtering. Our experiments consist of 30,000 kick-and-save tests, using 100 random angle and distance kicks from six distance categories and four angle categories repeated 30 times. Our benchmark results show that both proposed approaches bring significant improvements for the goalkeeper’s save success rates ( \(>\) 200 %) and validate the applicability of the novel mental model based decision-making process. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Training to learn the use and maintenance of biomedical devices have various constraints. In order to complete these trainings, we proposed to use virtual reality based on pedagogical scenarios and Intelligent Tutoring Systems (ITS). In this paper, we first established the existing pedagogical scenario models and ITS. Subsequently we presented our proposal of a formal model based on the concept of learning organization by extension of UML in order to describre some pedagogical scenario and ITS. The use of this model is illustrated by an application of a virtual biomedical analyzer with the aim of learning the technical procedures of the device. Finally, we performed two experiments in order to verify the efficiency of virtual reality training.

ABSTRACT To be believable, virtual entities must be equipped with the ability to anticipate, that is, to predict the behavior of other entities and the subsequent consequences on the environment. For that purpose, we propose an original approach where the entity possesses an autonomous world of simulation within simulation, in which it can simulate itself (with its own model of behavior) and simulate the environment (with the representation of the behaviors of the other entities). This principle is illustrated by the development of an artificial juggler in 3D. In this application, the juggler predicts the motion of the balls in the air and uses its predictions to coordinate its own behavior to continue to juggle.Copyright © 2012 John Wiley & Sons, Ltd.

ABSTRACT In some video games, humans and computer programs can play together, each one controlling a virtual humanoid. These computer programs usually aim at replacing missing human players; however, they partially miss their goal, as they can be easily spotted by players as being artificial. Our objective is to find a method to create programs whose behaviors cannot be told apart from players when observed playing the game. We call this kind of behavior abelievable behavior. To achieve this goal, we choose models using Markov chains to generate the behaviors by imitation. Such models use probability distributions to find which decision to choose depending on the perceptions of the virtual humanoid. Then, actions are chosen depending on the perceptions and the decision. We propose a new model, calledChameleon, to enhance expressiveness and the associated imitation learning algorithm. We first organize the sensors and motors by semantic refinement and add a focus mechanism in order to improve the believability. Then, we integrate an algorithm to learn the topology of the environment that tries to best represent the use of the environment by the players. Finally, we propose an algorithm to learn parameters of the decision model. Copyright © 2013 John Wiley & Sons, Ltd.

Abstract This paper deals with simulations of real‐time interactive character behavior. The underlying idea is to take into account principles from cognitive science, in particular, the human ability to anticipate and simulate the world behavior. For that purpose, we propose a conceptual framework where the entity possesses an autonomous world of simulation within simulation, in which it can simulate itself (with its own model of behavior) and the environment (with an abstract representation, which can be learnt, of the other entities behaviors). This principle is illustrated by the development of an artificial juggler, which predicts the motion of balls in the air and uses its predictions to coordinate its own behavior while juggling. Thanks to this model it is possible to add a human user to launch balls that the virtual juggler can catch whilst juggling. Copyright © 2011 John Wiley & Sons, Ltd.

Abstract This paper focuses on the simulation of behavior for autonomous entities in virtual environments. The behavior of these entities must determine their responses not only to external stimuli, but also with regard to internal states. We propose to describe such behavior using fuzzy cognitive maps (FCMs), whereby these internal states might be explicitly represented. This paper presents the use of FCMs as a tool to specify and control the behavior of individual agents. First, we describe how FCMs can be used to model behavior. We then present a learning algorithm allowing the adaptation of FCMs through observation. Copyright © 2010 John Wiley & Sons, Ltd.

This paper presents an educational distributed virtual reality-based environment for children called EVE-Environnement Virtuel pour Enfants. EVE is used in elementary schools from different countries as a supplementary tool in teaching children to read. This virtual environment supports cooperation among members of a dispersed team engaged in a shared context. By the mean of their avatars, which are special cases of agents, children are allowed to interact and to give decisions using cooperative mechanisms. The virtual environment architecture is reactive agent-based. The FCM-like dynamic action planning mechanism assures the agent's adaptability to its environmental changes. It also allows the pedagogical agent to adapt its behaviour to the child's actions. The implementation is based on a client-server architecture, VRML and C++ as languages and AReVi as graphical rendering API.

This study concerns virtual environments for training in operational conditions. The principal developed idea is that these environments are heterogeneous and open multi-agent systems. The MASCARET model is proposed to organize the interactions between agents and to provide them reactive, cognitive and social abilities to simulate the physical and social environment. The physical environment represents, in a realistic way, the phenomena that learners and teachers have to take into account. The social environment is simulated by agents executing collaborative and adaptive tasks. These agents realize, in team, procedures that they have to adapt to the environment. The users participate to the training environment through their avatar. In this article, we explain how we integrated in MASCARET models necessary to the creation of Intelligent Tutoring System. We notably incorporate pedagogical strategies and pedagogical actions. We present pedagogical agents. To validate our model, the SECUREVI application for fire-fighters training is developed.