We present the Robot Bookworm, a multi-session intervention co-designed with children and educators to foster reading motivation through personalized book discussions. The robot assigned each child a personally fitting book and engaged them in pedagogically structured discussions, with personalized book-aligned dialogic content selectively generated offline by a language model and moderated by people to ensure safety. We compared a personalized book discussion condition with a book-neutral control in a four-session, large-scale user study in two primary schools (N = 101, 8-11 y.o.). The intervention significantly increased reader-book relatedness and reading enjoyment, particularly for children with below-ceiling baseline enjoyment, but had no effect on intrinsic motivation. At a one-year follow-up, the quantitative effects were not sustained. However, children reported perceived positive shifts in attitudes towards reading, which they attributed to the Robot Bookworm.

LLM-based agents have emerged as transformative tools capable of executing complex tasks through iterative planning and action, achieving significant advancements in understanding and addressing user needs. Yet, their effectiveness remains limited in specialized domains such as mental health diagnosis, where they underperform compared to general applications. Current approaches to integrating diagnostic capabilities into LLMs rely on scarce, highly sensitive mental health datasets, which are challenging to acquire. These methods also fail to emulate clinicians' proactive inquiry skills, lack multi-turn conversational comprehension, and struggle to align outputs with expert clinical reasoning. To address these gaps, we propose DSM5AgentFlow, the first LLM-based agent workflow designed to autonomously generate DSM-5 Level-1 diagnostic questionnaires. By simulating therapist-client dialogues with specific client profiles, the framework delivers transparent, step-by-step disorder predictions, producing explainable and trustworthy results. This workflow serves as a complementary tool for mental health diagnosis, ensuring adherence to ethical and legal standards. Through comprehensive experiments, we evaluate leading LLMs across three critical dimensions: conversational realism, diagnostic accuracy, and explainability. Our datasets and implementations are fully open-sourced.

Learning from Demonstrations (LfD) allows robots to learn skills from human users, but its effectiveness can suffer due to sub-optimal teaching, especially from untrained demonstrators. Active LfD aims to improve this by letting robots actively request demonstrations to enhance learning. However, this may lead to frequent context switches between various task situations, increasing the human cognitive load and introducing errors to demonstrations. Moreover, few prior studies in active LfD have examined how these active query strategies may impact human teaching in aspects beyond user experience, which can be crucial for developing algorithms that benefit both robot learning and human teaching. To tackle these challenges, we propose an active LfD method that optimizes the query sequence of online human demonstrations via Curriculum Learning (CL), where demonstrators are guided to provide demonstrations in situations of gradually increasing difficulty. We evaluate our method across four simulated robotic tasks with sparse rewards and conduct a user study$(N=26)$to investigate the influence of active LfD methods on human teaching regarding teaching performance, post-guidance teaching adaptivity, and teaching transferability. Our results show that our method significantly improves learning performance compared to three other LfD baselines in terms of the final success rate of the converged policy and sample efficiency. Additionally, results from our user study indicate that our method significantly reduces the time required from human demonstrators and decreases failed demonstration attempts. It also enhances post-guidance human teaching in both seen and unseen scenarios compared to another active LfD baseline, indicating enhanced teaching performance, greater postguidance teaching adaptivity, and better teaching transferability achieved by our method.

When encountering a robot in the wild, it is not inherently clear to human users what the robot's capabilities are. When encountering misunderstandings or problems in spoken interaction, robots often just apologize and move on, without additional effort to make sure the user understands what happened. We set out to compare the effect of two speech based capability communication strategies (proactive, reactive) to a robot without such a strategy, in regard to the user's rating of and their behavior during the interaction. For this, we conducted an in-person user study with 120 participants who had three speech-based interactions with a social robot in a restaurant setting. Our results suggest that users preferred the robot communicating its capabilities proactively and adjusted their behavior in those interactions, using a more conversational interaction style while also enjoying the interaction more.

With large language models (LLMs) on the rise, in-game interactions are shifting from rigid commands to natural conversations. However, the impacts of LLMs on player performance and game experience remain underexplored. This work explores LLM's role as a co-builder during gameplay, examining its impact on task performance, usability, and player experience. Using Minecraft as a sandbox, we present an LLM-assisted interface that engages players through natural language, aiming to facilitate creativity and simplify complex gaming commands. We conducted a mixed-methods study with 30 participants, comparing LLM-assisted and command-based interfaces across simple and complex game tasks. Quantitative and qualitative analyses reveal that the LLM-assisted interface significantly improves player performance, engagement, and overall game experience. Additionally, task complexity has a notable effect on player performance and experience across both interfaces. Our findings highlight the potential of LLM-assisted interfaces to revolutionize virtual experiences, emphasizing the importance of balancing intuitiveness with predictability, transparency, and user agency in AI-driven, multimodal gaming environments.

Transfer Learning (TL) is a powerful tool that enables robots to transfer learned policies across different environments, tasks, or embodiments. To further facilitate this process, efforts have been made to combine it with Learning from Demonstrations (LfD) for more flexible and efficient policy transfer. However, these approaches are almost exclusively limited to offline demonstrations collected before policy transfer starts, which may suffer from the intrinsic issue of covariance shift brought by LfD and harm the performance of policy transfer. Meanwhile, extensive work in the learning-from-scratch setting has shown that online demonstrations can effectively alleviate covariance shift and lead to better policy performance with improved sample efficiency. This work combines these insights to introduce online demonstrations into a policy transfer setting. We present Policy Transfer with Online Demonstrations, an active LfD algorithm for policy transfer that can optimize the timing and content of queries for online episodic expert demonstrations under a limited demonstration budget. We evaluate our method in eight robotic scenarios, involving policy transfer across diverse environment characteristics, task objectives, and robotic embodiments, with the aim to transfer a trained policy from a source task to a related but different target task. The results show that our method significantly outperforms all baselines in terms of average success rate and sample efficiency, compared to two canonical LfD methods with offline demonstrations and one active LfD method with online demonstrations. Additionally, we conduct preliminary sim-to-real tests of the transferred policy on three transfer scenarios in the real-world environment, demonstrating the policy effectiveness on a real robot manipulator.

The present study aims at investigating how humans engage with common communication modalities—speech, tablet, and gesture—when interacting with a humanoid robot. To explore this, we designed a live interaction experiment using a congruence paradigm, where participants engaged with a robot presenting two out of three modalities simultaneously: one as the primary cue and the other as a distracting cue. We measured participants’ task performance (response time, error rate) and fixation distribution (fixation count and duration proportions) across different roles (primary, distracting, neither) and areas of interest (face, tablet, gesture). Additionally, we compared fixation patterns between the performance and baseline phases. Our findings reveal that while the tablet is the most effective modality for task engagement, it also serves as a strong attentional distractor, dominating gaze allocation regardless of its informational value. This underscores the importance of carefully balancing tablet integration in HRI design. Notably, our results demonstrate that gaze patterns alone do not fully reveal attentional focus, emphasizing the need to consider both overt and covert cognitive processes in multimodal HRI. These insights provide valuable guidelines for designing more effective and engaging human-robot interactions.

Speech recognition in Human-Robot Interaction (HRI) fully relies on audio-based Automatic Speech Recognition. However, speech recognition that relies solely on audio faces significant challenges in noisy environments and may lead to poor performance in such environments. One approach to address this is to also use lipreading in combination with traditional speech recognition. Recent work has shown that audiovisual speech recognition (AVSR) can achieve a Word Error Rate (WER) of only 0.9% on the dataset LRS3. In this paper, we assess the potential of combining audio with lipreading on a social robot platform, Pepper, which has not yet been widely tested for AVSR. Given that prior research has focused on non-robotic domains, it remains unclear whether such models can generalize well to social robot environments. We systematically evaluate and compare the performance of established offline and real-time audiovisual models with their audio-only counterparts. The experiments were conducted in both a controlled laboratory setting and a dynamic and noisy public environment. We evaluated the data using WER and also measured the inference latency of real-time models via Real-Time Factor and Words Per Second rates. The results demonstrate real-time performance for audio-only speech recognition across all latency metrics and near real-time performance for models that combine audio with lipreading. We also explored factors that might influence the inference performance of these models to understand how much video contributes to the audio. This includes factors related to (1) environmental and temporal variations, (2) model behavior, and (3) implementation choices. Our findings indicate that for now the audio-only models outperform the audiovisual models on a social robot platform, in contrast to what has been reported in the benchmarked literature. We conclude that more work is still needed to benefit from lipreading in HRI.

Appearance-based gaze estimation, an accessible and unobtrusive alternative to eye tracking, has advanced significantly, yet their adoption in human-robot interaction (HRI) remains limited. A key barrier is the lack of clarity on how they compare to high-precision eye trackers. To address this, we evaluate this method against eye-tracker glasses in an HRI setting using calibration and attention detection tasks. We assess performance across different cameras (4K and robot’s built-in camera) and participant conditions (with and without glasses). Results show that the 4K camera and participants without glasses yield higher accuracy and precision. With a simple offset correction, this method achieves comparable performance to eye-tracker glasses for average gaze pattern but struggles with detecting gaze patterns over time. It also demonstrates potential for real-time robot attention detection. We conclude that appearance-based gaze estimation is a viable, cost-effective alternative to traditional eye tracking in HRI, particularly for average gaze pattern detection.

Reinforcement Learning (RL) has achieved great success in sequential decision-making problems but often requires extensive agent-environment interactions. To improve sample efficiency, methods like Reinforcement Learning from Expert Demonstrations (RLED) incorporate external expert demonstrations to aid agent exploration during the learning process. However, these demonstrations, typically collected from human users, are costly and thus often limited in quantity. Therefore, how to select the optimal set of human demonstrations that most effectively aids learning becomes a critical concern. This paper introduces EARLY (Episodic Active Learning from demonstration querY), an algorithm designed to enable a learning agent to generate optimized queries for expert demonstrations in a trajectory-based feature space. EARLY employs a trajectory-level estimate of uncertainty in the agent’s current policy to determine the optimal timing and content for feature-based queries. By querying episodic demonstrations instead of isolated state-action pairs, EARLY enhances the human teaching experience and achieves better learning performance. We validate the effectiveness of our method across three simulated navigation tasks of increasing difficulty. Results indicate that our method achieves expert-level performance in all three tasks, converging over 50% faster than other four baseline methods when demonstrations are generated by simulated oracle policies. A follow-up pilot user study (N = 18) further supports that our method maintains significantly better convergence with human expert demonstrators, while also providing a better user experience in terms of perceived task load and requiring significantly less human time.

Maintaining the child-robot relationship after a significant break, such as a holiday, is an important step for developing sustainable social robots for education. We ran a four-session user study (n = 113 children) that included a nine-month break between the third and fourth session. During the study, participants practiced math with the help of a social robot math tutor. We found that social personalization is an effective strategy to better sustain the child-robot relationship than the absence of social personalization. To become reacquainted after the long break, the robot summarizes a few pieces of information it had stored about the child. This gives children a feeling of being remembered, which is a key contributor to the effectiveness of social personalization. Enabling the robot to refer to information previously shared by the child is another key contributor to social personalization. Conditional for its effectiveness, however, is that children notice these memory references. Finally, although we found that children's interest in the tutoring content is related to relationship formation, personalizing the topics did not lead to more interest in the content. It seems likely that not all of the memory information that was used to personalize the content was up-to-date or socially relevant.

Identifying gaze targets in videos of human-robot interaction is useful for measuring engagement. In practice, this requires manually annotating for a fixed set of objects that a participant is looking at in a video, which is very time-consuming. To address this issue, we propose an annotation pipeline for automating this effort. In this work, we focus on videos in which the objects looked at do not move. As input for the proposed pipeline, we therefore only need to annotate object bounding boxes for the first frame of each video. The benefit, moreover, of manually annotating these frames is that we can also draw bounding boxes for objects outside of it, which enables estimating gaze targets in videos where not all objects are visible. A second issue that we address is that the models used for automating the pipeline annotate individual video frames. In practice, however, manual annotation is done at the event level for video segments instead of single frames. Therefore, we also introduce and investigate several variants of algorithms for aggregating frame-level to event-level annotations, which are used in the last step in our annotation pipeline.We compare two versions of our pipeline: one that uses a state-of-the-art gaze estimation model (GEM) and a second one using a state-of-the-art target detection model (TDM). Our results show that both versions successfully automate the annotation, but the GEM pipeline performs slightly (≈10%) better for videos where not all objects are visible. Analysis of our aggregation algorithm, moreover, shows that there is no need for manual video segmentation because a fixed time interval for segmentation yields very similar results. We conclude that the proposed pipeline can be used to automate almost all of the annotation effort.

Recent models for Visual Speech Recognition (VSR) have shown remarkable progress over the last few years. They have however been applied mainly to datasets such as Lip Reading Sentences 3 (LRS3), LRS2 or Lombard GRID, but not yet on social robots. As social robots struggle to recognize speech in more challenging acoustic and crowded environments, we believe such models are promising tools for real-time interaction with users. This paper presents a feasibility study focusing on integration of speech recognition (SR) using mixed modalities - audio, visual (lip-reading) and audio-visual - in social robots. To this end, this paper contributes a pipeline to detect an active speaker based on lip movement, post-processing of audio and video footage and inferencing it with the state-of-the-art Auto-AVSR model. In a user study (N = 26), we evaluated the feasibility of audio, visual and mixed modality speech recognition on a Pepper robot. We demonstrate the feasibility of using singular and mixed modalities with speech-to-text inference in natural interaction. The results show that it is feasible to deploy such models on social robots in a controlled, noiseless and non-interactive environment. Additionally, the results revealed that informing participants to emphasize their lip movements significantly improved text-to-speech inference results. Our work provides initial insights into the benefits and challenges of using VSR, ASR and AVSR for HRI.

With current state-of-the-art (SOTA) automatic speech recognition (ASR) systems, it is not possible to transcribe overlapping speech audio streams separately. Consequently, when these ASR systems are used as part of a social robot like Pepper for interaction with a human, it is common practice to close the robot’s microphone while it is talking itself. This prevents the human users to interrupt the robot, which limits speech-based human-robot interaction. To enable a more natural interaction which allows for such interruptions, we propose an audio processing pipeline for filtering out robot’s ego speech using only a single-channel microphone. This pipeline takes advantage of the possibility to feed the robot ego speech signal, generated by a text-to-speech API, as training data into a machine learning model. The proposed pipeline combines a convolutional neural network and spectral subtraction to extract overlapping human speech from the audio recorded by the robot-embedded microphone. When evaluating on a held-out test set, we find that this pipeline outperforms our previous approach to this task, as well as SOTA target speech extraction systems that were retrained on the same dataset. We have also integrated the proposed pipeline into a lightweight robot software development framework to make it available for broader use. As a step towards demonstrating the feasibility of deploying our pipeline, we use this framework to evaluate the effectiveness of the pipeline in a small lab-based feasibility pilot using the social robot Pepper. Our results show that when participants interrupt the robot, the pipeline can extract the participant’s speech from one-second streaming audio buffers received by the robot-embedded single-channel microphone, hence in near-real time.

As social robots see increasing deployment within the general public, improving the interaction with those robots is essential. Spoken language offers an intuitive interface for the human–robot interaction (HRI), with dialogue management (DM) being a key component in those interactive systems. Yet, to overcome current challenges and manage smooth, informative, and engaging interaction, a more structural approach to combining HRI and DM is needed. In this systematic review, we analyze the current use of DM in HRI and focus on the type of dialogue manager used, its capabilities, evaluation methods, and the challenges specific to DM in HRI. We identify the challenges and current scientific frontier related to the DM approach, interaction domain, robot appearance, physical situatedness, and multimodality.

To benefit from the social capabilities of a robot math tutor, instead of being distracted by them, a novel approach is needed where the math task and the robot's social behaviors are better intertwined. We present concrete design specifications of how children can practice math via a personal conversation with a social robot and how the robot can scaffold instructions. We evaluated the designs with a three-session experimental user study (n = 130, 8-11 y.o.). Participants got better at math over time when the robot scaffolded instructions. Furthermore, the robot felt more as a friend when it personalized the conversation.

In order to improve human-agent interaction, it is essential to have good measures of interaction quality. We define interaction quality based on multiple aspects, including usability, likability and perceived conversation quality as subjective measures, and interaction length, completion rate and frequency of unrecognized utterances as objective measures. Determining necessary improvements to a conversational agent is a non-trivial task, because it is difficult to infer from an evaluation of the agent as a whole, which aspects of the agent need to be improved to raise the interaction quality. In this paper, we propose a scoring system for task-oriented conversational agents to predict aspects of interaction quality and to guide an iterative improvement process. Our scoring system does not provide a single score, but leverages structural features of the dialogue management approach and assigns a score on three levels: the utterance, dialogue move, and genre level. Using the agent's scores on separate levels to predict the interaction quality allows making targeted improvements to the conversational agent. In order to evaluate our scoring system, we apply it over the course of multiple crowdsourcing pilot studies, using a recipe recommendation agent. We evaluate the obtained scores in regard to their ability to predict selected objective and subjective interaction quality aspects, as well as their suitability for making informed decisions about necessary improvements.

Appearance-based methods are a promising solution for gaze estimation, as they eliminate the need for additional devices and calibration. This makes them particularly well-suited for human-robot interaction (HRI) research. However, until recently their performance was under par compared to traditional eye-trackers. Recent breakthroughs have been made with the release of two large-scale datasets with a wide range of gaze directions (Gaze360 and ETH-XGaze) and the accompanying state-of-the-art deep neural networks (L2CS and ETH). In this paper, we systematically evaluate the performance of these two appearance-based models on a social robot. In our setup, we vary the distance from the robot (1-3m) and camera resolution (640*480 and 3840*2160) and analyze the performance in terms of accuracy and precision. We find that the L2CS model trained on the Gaze360 dataset combined with a 4K camera achieves the best performance on the 2 m and 3 m distances. We show that a simple offset correction on pitch and yaw can further increase the accuracy and precision by 18.6% and 9.6% respectively. We conclude that for a range up to 3 m appearance-based gaze estimation models provide a promising approach for application in HRI research.

Learning from Demonstrations (LfD) transfers skills from human teachers to robots. However, data imbalance in demonstrations can bias policies towards majority situations. Previous work attempted to solve this problem after data collection, but few efforts were made to maintain a balanced distribution from the phase of data acquisition. Our method accounts for the influence of robots on human teachers and enables robots to actively guide interaction to approximate demonstration distributions to target distributions. Simulated and real-world experiments validated the method’s efficacy in shaping demonstration distribution into various target distributions and robustness to various levels of uncertainties. Also, our method significantly improved the generalization ability of robot learning when LfD policies were trained with data collected by our method compared to natural data collection.

Human-centered interactive robot tasks (e.g., social greetings and cooperative dressing) are a type of task where humans are involved in task dynamics and performance evaluation. Such tasks require spatial and temporal coordination between agents in real-time, tackling physical limitations from constrained robot bodies, and connecting human user experience with concrete learning objectives to inform algorithm design. To solve these challenges, imitation learning has become a popular approach where by a robot learns to perform a task by imitating how human experts do it (i.e., expert policies). However, previous works tend to isolate the algorithm design from the design of the whole learning pipeline, neglecting its connection with other modules inside the process (like data collection and user-centered subjective evaluation) from the view as a system. Going beyond traditional imitation learning, this work reexamines robot imitation learning in human-centered interactive tasks from the perspective of the whole learning pipeline, ranging from data collection to subjective evaluation. We present a process-oriented framework that consists of a guideline to collect diverse yet representative demonstrations and an interpreter to explain subjective user-centered performance with objective robot-related parameters. We illustrate the steps covered by the framework in a fist-bump greeting task as demonstrative deployment. Results show that our framework is able to identify representative human-centered features to instruct demonstration collection and validate influential robot-centered factors to interpret the gap in subjective performance between the expert policy and the imitator policy.

It is useful for a social robot to detect and locate users based on their speech. Notable challenges hampering the effective localization of a speaker are background noise and overlapping speech. Convolutional Neural Networks (CNNs) have shown to yield good performance on locating single speakers on a curated dataset, but to a lesser extent in scenarios with two speakers. In addition, their computational cost is still too high for a timely reaction in real-world settings. We build on the current state-of-the-art CNN approach, and propose several improvements for distinguishing multiple speakers by time-alignment in the input representation and reducing computational costs by considerably shortening the input audio blocks. We evaluate this approach on an existing dataset with blocks of noisy and overlapping speech recorded in rooms of different sizes, predicting the number of active speech events and their azimuth locations. The results show that our approach outperforms other approaches in locating two speakers and is considerably faster than the best-performing alternative approach. The time-domain information in the input representation was found essential for predicting the location of the signal source.

Abstract Emotions are known to spread among people, a process known as emotion contagion. Both positive and negative emotions are believed to be contagious, but the mass spread of negative emotions has attracted the most attention due to its danger to society. The use of agent-based techniques to simulate emotion contagion in crowds has grown over the last decade and a range of contagion mechanisms and applications have been considered. With this review we aim to give a comprehensive overview of agent-based methods to implement emotion contagion in crowd simulations. We took a systematic approach and collected studies from Web of Science, Scopus, IEEE and ACM that propose agent-based models that include a process of emotion contagion in crowds. We classify the models in three categories based on the mechanism of emotion contagion and analyse the contagion mechanism, application and findings of the studies. Additionally, a broad overview is given of other agent characteristics that are commonly considered in the models. We conclude that there are fundamental theoretical differences among the mechanisms of emotion contagion that reflect a difference in view on the contagion process and its application, although findings from comparative studies are inconclusive. Further, while large theoretical progress has been made in recent years, empirical evaluation of the proposed models is lagging behind due to the complexity of reliably measuring emotions and context in large groups. We make several suggestions on a way forward regarding validation to eventually justify the application of models of emotion contagion in society.

This article aims to summarize and describe research on the topic of automatic group emotion recognition. In recent years, the topic of emotion analysis of groups or crowds has gained interest, with studies performing emotion detection in different contexts, using different datasets and modalities (such as images, video, audio, social media messages), and taking different approaches. Articles are included after an innovative search method, including Dense Query Extraction and automatic cross-referencing. Discussed are the types of groups and emotion models considered in automatic emotion recognition research, common datasets for all modalities, general approaches taken, and reported performances. These performances are discussed, followed by an analysis of the application possibilities of the discussed methods. To ensure clear, replicable, and comparable studies, we suggest research should test on multiple, common datasets and report on multiple metrics, when possible. Implementation details and code should be made available where possible. An area of interest for future work is to build systems with more real-world application possibilities, coping with changing group sizes, different emotional subgroups, and changing emotions over time, while having a higher robustness and working with datasets with reduced biases.

While interacting with a social robot, children have a need to express themselves and have their expressions acknowledged by the robot—a need that is often unaddressed by the robot, due to its limitations in understanding the expressions of children. To keep the child-robot interaction manageable, the robot takes control, undermining children’s ability to co-regulate the interaction. Co-regulation is important for having a fulfilling social interaction. We developed a co-creation activity that aims to facilitate more co-regulation. Children are enabled to create sound effects, gestures, and light animations for the robot to use during their conversation. A crucial additional feature is that children are able to coordinate their involvement of the co-creation process. Results from a user study (n= 59 school children, 7–11 years old) showed that the co-creation activity successfully facilitated co-regulation by improving children’s agency. It also positively affected the acceptance of the robot. We furthermore identified five distinct profiles detailing the different needs and motivations children have for the level of involvement they chose during the co-creation process.

After the novelty effect wears off children need a new motivator to keep interacting with a social robot. Enabling children to build a relationship with the robot is the key for facilitating a sustainable long-term interaction. We designed a memory-based personalization strategy that safeguards the continuity between sessions and tailors the interaction to the child's needs and interests to foster the child-robot relationship. A longitudinal (five sessions in two months) user study (N = 46, 8–10 y.o) showed that the strategy kept children interested longer in the robot, fosters more closeness, elicits more positive social cues, and adds continuity between sessions.

A key challenge in human-robot interaction (HRI) design is to create and sustain engaging social interactions. This paper argues that improvisational techniques from the performing arts can address this challenge. Contrary to the ways in which improvisation is generally used in social robotics, we propose an understanding of improvisational techniques as based on rules that shape motion choices. We claim that such an approach, represented in what we name the “external” and “emergent” perspectives on improvisation, could benefit the way in which robot movement and behaviour is designed and deployed, increasing playful engagement and responsiveness. As an example of this type of improvisation, we discuss how American dancer and choreographer William Forsythe's Improvisation Technologies could be used in an HRI context. We also report on a preliminary experimentation using a Wizard-of-Oz exploratory prototyping system and a participatory design method with professional dancers geared towards the exploration of interactive movement possibilities with a Pepperrobot. Finally, we report on how this workshop offered valuable information about the applicability of these tools, as well as reflections on how it could help increase the level of engagement in the interaction.

Both social group detection and group emotion recognition in images are growing fields of interest, but never before have they been combined. In this work we aim to detect emotional subgroups in images, which can be of great importance for crowd surveillance or event analysis. To this end, human annotators are instructed to label a set of 171 images, and their recognition strategies are analysed. Three main strategies for labeling images are identified, with each strategy assigning either 1) more weight to emotions (emotion-based fusion), 2) more weight to spatial structures (group-based fusion), or 3) equal weight to both (summation strategy). Based on these strategies, algorithms are developed to automatically recognize emotional subgroups. In particular, K-means and hierarchical clustering are used with location and emotion features derived from a fine-tuned VGG network. Additionally, we experiment with face size and gaze direction as extra input features. The best performance comes from hierarchical clustering with emotion, location and gaze direction as input.

Social robots have been introduced in different fields such as retail, health care and education. Primary education in the Netherlands (and elsewhere) recently faced new challenges because of the COVID-19 pandemic, lockdowns and quarantines including students falling behind and teachers burdened with high workloads. Together with two Dutch municipalities and nine primary schools we are exploring the long-term use of social robots to study how social robots might support teachers in primary education, with a focus on mathematics education. This paper presents an explorative study to define requirements for a social robot math tutor. Multiple focus groups were held with the two main stakeholders, namely teachers and students. During the focus groups the aim was 1) to understand the current situation of mathematics education in the upper primary school level, 2) to identify the problems that teachers and students encounter in mathematics education, and 3) to identify opportunities for deploying a social robot math tutor in primary education from the perspective of both the teachers and students. The results inform the development of social robots and opportunities for pedagogical methods used in math teaching, child-robot interaction and potential support for teachers in the classroom.

Clothing is often used as a way to communicate about one’s identity. It can provide information about characteristics such as gender, age, occupation, and status as well as more subjective attributes such as trustworthiness and friendliness. In this paper we report on exploratory research that we conducted by means of three studies to investigate the effects of clothing on a Pepper robot. Three outfits that varied in style and colour were compared with each other and the robot without clothing. Findings from an online survey suggest that gender perception but not human-likeness can be manipulated by clothing. We also found that first impressions on expertise and likeability may vary with clothing style and may induce stereotypical job associations. In a second study, we interviewed experts which were all familiar with the Pepper platform to obtain a more qualitative perspective on robot clothing. The interviews made clear that specific features of clothing may trigger strong associations with, for example, occupations, and that features such as headdress may make an outfit look more like a uniform. Finally, we conducted a field experiment comparing two clothing conditions for a receptionist robot in a natural setting and found that a robot with uniform appears to be more engaging. Our findings suggest that robot clothing may have an effect on first impressions, gender perception, and engagement in interactions with users in the wild.

Demonstrating reliability of cognitive multi-agent systems is of key importance. There has been an extensive amount of work on logics for verifying cognitive agents but it has remained mostly theoretical. Cognitive agent-oriented programming languages provide the tools for compact representation of complex decision making mechanisms, which offers an opportunity for applying a theorem proving approach. We base our work on the belief that theorem proving can add to the currently available approaches for providing assurance for cognitive multi-agent systems. However, a practical approach using theorem proving is missing. We explore the use of proof assistants to make verifying cognitive multi-agent systems more practical.

In this paper we specify and validate three interaction design patterns for an interactive storytelling experience with an autonomous social robot. The patterns enable the child to make decisions about the story by talking with the robot, reenact parts of the story together with the robot, and recording self-made sound effects. The design patterns successfully support children's engagement and agency. A user study (N = 27, 8-10 y.o.) showed that children paid more attention to the robot, enjoyed the storytelling experience more, and could recall more about the story, when the design patterns were employed by the robot during storytelling. All three aspects are important features of engagement. Children felt more autonomous during storytelling with the design patterns and highly appreciated that the design patterns allowed them to express themselves more freely. Both aspects are important features of children's agency. Important lessons we have learned are that reducing points of confusion and giving the children more time to make themselves heard by the robot will improve the patterns efficiency to support engagement and agency. Allowing children to pick and choose from a diverse set of stories and interaction settings would make the storytelling experience more inclusive for a broader range of children.

Emotion expression is an important part of human-robot interaction. Previous studies typically focused on a small set of emotions and a single channel to express them. We developed an emotion expression model that modulates motion, poses and LED features parametrically, using valence and arousal values. This model does not interrupt the task or gesture being performed and hence can be used in combination with functional behavioural expressions. Even though our model is relatively simple, it is just as capable of expressing emotions as other more complicated models that have been proposed in the literature. We systematically explored the expressivity of our model and found that a parametric model using 5 key motion and pose features can be used to effectively express emotions in the two quadrants where valence and arousal have the same sign. As paradigmatic examples, we tested for happy, excited, sad and tired. By adding a second channel (eye LEDs), the model is also able to express high arousal (anger) and low arousal (relaxed) emotions in the two other quadrants. Our work supports other findings that it remains hard to express moderate arousal emotions in these quadrants for both negative (fear) and positive (content) valence.

We design and evaluate a robot interviewer for collecting visitor data in a museum for marketing purposes. We take inspiration from research on face-to-face human intercept interviews. We develop a personal interviewing style that is expected to motivate participants to answer more questions and compare this with a more formal style. We also evaluate whether a greeting ritual performed by the robot increases participation and whether taking a picture with the robot is an effective incentive for visitors to participate in an interview.Our study is conducted "in the wild" and we analyse sessions with the robot and passersby in a museum. The independent variables were interviewing style and whether an incentive was offered or not. The dependent variables were participation and continuation rate, and museum ratings. Contrary to expectations, we find that the participation rate is lower when the robot provides an incentive. Although we find that a personal style is perceived as more social, it does not influence the continuation rate. Museum ratings were also not affected by style. Our style manipulation may not have been strong enough to produce these effects.Our study shows that social robots have a high potential for conducting intercept interviews. Willingness to participate in a robot interview is high, while this is one of the main challenges with intercept interviews. To improve data collection, people detection and speech recognition skills could be improved.

Explanation of actions is important for transparency of-, and trust in the decisions of smart systems. Literature suggests that emotions and emotion words - in addition to beliefs and goals - are used in human explanations of behaviour. Furthermore, research in e-health support systems and human-robot interaction stresses the need for studying long-term interaction with users. However, state of the art explainable artificial intelligence for intelligent agents focuses mainly on explaining an agent's behaviour based on the underlying beliefs and goals in short-term experiments. In this paper, we report on a long-term experiment in which we tested the effect of cognitive, affective and lack of explanations on children's motivation to use an e-health support system. Children (aged 6-14) suffering from type 1 diabetes mellitus interacted with a virtual robot as part of the e-health system over a period of 2.5 - 3 months. Children alternated between the three conditions. Agent behaviours that were explained to the children included why 1) the agent asks a certain quiz question; 2) the agent provides a specific tip (a short instruction) about diabetes; or, 3) the agent provides a task suggestion, e.g., play a quiz, or, watch a video about diabetes. Their motivation was measured by counting how often children would follow the agent's suggestion, how often they would continue to play the quiz or ask for an additional tip, and how often they would request an explanation from the system. Surprisingly, children proved to follow task suggestions more often when no explanation was given, while other explanation effects did not appear. This is to our knowledge the first longterm study to report empirical evidence for an agent explanation effect, challenging the next studies to uncover the underlying mechanism.

Nowadays, off-the-shelf social robots are used more frequently by the HRI community to research social interactions with different types of users across a range of domains such as education, retail, health care, public places and other domains. Everyone doing HRI research with end-users is invited to submit a case study to our workshop. We are particularly interested in case studies where things did not go as planned. Case studies describing research in the lab or in the wild are both welcome. Examples of unplanned experiences could include, but are not limited to, unexpected responses from the user, issues with the experimental setup or simply having challenges with transferring theory to the real world. In this workshop, we focus on off-the-shelf robots. In order to generalize and compare differences across multiple HRI domains and create common solutions, we will provide a template for your case study. We are interested in learning how such unexpected HRI results can be reported. In the workshop, we will discuss and study how failures are reported and be inspired to create a list of good ways to report failures, which can hopefully be inspiring for the HRI community.

Humans use several social cues, both verbal and nonverbal, to draw the attention of others. In this study we investigate whether similar behaviors can also be effectively used by a social robot for drawing attention. To this end, we setup a welcoming humanoid (Pepper) at the entrance of a university building. Its behaviors include one or a combination of behavioral modalities (i.e., a waving gesture, utterance and movement). These behaviors are triggered automatically based on people detection software which tracks passersby and monitors their head keypoints. Our findings imply that Pepper draws more attention when displaying a combination of modalities.

Drawing the attention of passersby is a basic task of a social robot to initiate an interaction in a public environment (e.g., shopping malls, museums or hospitals). Humans use several social cues, both verbal and nonverbal, to draw the attention of others. In this study, we investigate whether similar behaviors can also be effectively used by a social robot for drawing attention. To this end, we setup a humanoid robot (Pepper) to act as a welcoming robot at the entrance of a university building. The behaviors selected for Pepper include one or a combination of behavioral modalities (i.e., a waving gesture, utterance and movement). These behaviors are triggered automatically using the output of people detection software which tracks passersby and monitors their head keypoints (nose, eyes, and ears). The reactions of people toward Pepper are observed and recorded by means of an observation sheet. For several weeks, we deployed Pepper at the entrance with the aim of wearing off the novelty effect. In our final study, we collected data from several hundreds of passersby (N=364) and conducted post-interviews with randomly selected ones (N=28). Passersby noticed Pepper at the entrance and clearly recognized its role as a welcoming robot. In addition, Pepper was able to draw more attention when displaying a combination of behavioral modalities. However, passersby did not recall the robot utterance as they, for example, were unable to reproduce it or mistakenly claimed that the robot said something when it was only waving.

This paper presents the design and evaluation of human-like welcoming behaviors for a humanoid robot to draw the attention of passersby by following a three-step model: (1) selecting a target (person) to engage, (2) executing behaviors to draw the target's attention, and (3) monitoring the attentive response. A computer vision algorithm was developed to select the person, start the behaviors and monitor the response automatically. To vary the robot's enthusiasm when engaging passersby, a waving gesture was designed as basic welcoming behavioral element, which could be successively combined with an utterance and an approach movement. This way, three levels of enthusiasm were implemented: Mild (waving), moderate (waving and utterance) and high (waving, utterance and approach movement). The three levels of welcoming behaviors were tested with a Pepper robot at the entrance of a university building. We recorded data and observation sheets from several hundreds of passersby (N = 364) and conducted post-interviews with randomly selected passersby (N = 28). The level selection was done at random for each participant. The passersby indicated that they appreciated the robot at the entrance and clearly recognized its role as a welcoming robot. In addition, the robot proved to draw more attention when showing high enthusiasm (i.e., more welcoming behaviors), particularly for female passersby.

Exploration games are games where agents (or robots) need to search resources and retrieve these resources. In principle, performance in such games can be improved either by adding more agents or by exchanging more messages. However, both measures are not free of cost and it is important to be able to assess the trade-off between these costs and the potential performance gain. The focus of this paper is on improving our understanding of the performance gain that can be achieved either by adding more agents or by increasing the communication load. Performance gain moreover is studied by taking several other important factors into account such as environment topology and size, resource-redundancy, and task size. Our results suggest that there does not exist a decision function that dominates all other decision functions, i.e. is optimal for all conditions. Instead we find that (i) for different team sizes and communication strategies different agent decision functions perform optimal, and that (ii) optimality of decision functions also depends on environment and task parameters. We also find that it pays off to optimize for environment topologies. Copyright © 2018 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved. Institute for Systems and Technologies of Information, Control and Communication (INSTICC)

Task allocation and management is crucial for human-robot collaboration in Urban Search And Rescue response efforts. The job of a mission team leader in managing tasks becomes complicated when adding multiple and different types of robots to the team. Therefore, to effectively accomplish mission objectives, shared situation awareness and task management support are essential. In this paper, we design and evaluate an ontology which provides a common vocabulary between team members, both humans and robots. The ontology is used for facilitating data sharing and mission execution, and providing the required automated task management support. Relevant domain entities, tasks, and their relationships are modeled in an ontology based on vocabulary commonly used by firemen, and a user interface is designed to provide task tracking and monitoring. The ontology design and interface are deployed in a search and rescue system and its use is evaluated by firemen in a task allocation and management scenario. Results provide support that the proposed ontology (1) facilitates information sharing during missions; (2) assists the team leader in task allocation and management; and (3) provides automated support for managing an Urban Search and Rescue mission. Copyright © 2018 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved Institute for Systems and Technologies of Information, Control and Communication (INSTICC)

The annual International Automated Negotiating Agents Competition (ANAC) is used by the automated negotiation research community to benchmark and evaluate its work andto challenge itself. The benchmark problems and evaluation results and the protocols and strategies developed are available to the wider research community.

For real-time programs reproducing a bug by rerunning the system is likely to fail, making fault localization a time-consuming process. Omniscient debugging is a technique that stores each run in such a way that it supports going backwards in time. However, the overhead of existing omniscient debugging implementations for languages like Java is so large that it cannot be effectively used in practice. In this paper, we show that for agent-oriented programming practical omniscient debugging is possible. We design a tracing mechanism for efficiently storing and exploring agent program runs. We are the first to demonstrate that this mechanism does not affect program runs by empirically establishing that the same tests succeed or fail. Usability is supported by a trace visualization method aimed at more effectively locating faults in agent programs.

The main goal of our demonstration is to show how omniscient debugging can be applied in practice to cognitive agents. A concrete implementation of the mechanisms proposed in Koeman et. al [2017] has been created for the GOAL agent programming language in the Eclipse environment, integrated with the source-level debugger of Koeman et. al [2016], thus fully implementing the proposal within a state-of-the-art setting. The implementation will be used together with typical agent programs to demonstrate its practical use.

A good explanation takes the user who is receiving the explanation into account. We aim to get a better understanding of user preferences and the differences between children and adults who receive explanations from a robot. We implemented a Nao-robot as a belief-desire-intention (BDI)-based agent and explained its actions using two different explanation styles. Both are based on how humans explain and justify their actions to each other. One explanation style communicates the beliefs that give context information on why the agent performed the action. The other explanation style communicates the goals that inform the user of the agent's desired state when performing the action. We conducted a user study (19 children, 19 adults) in which a Nao-robot performed actions to support type 1 diabetes mellitus management. We investigated the preference of children and adults for goalversus belief-based action explanations. From this, we learned that adults have a significantly higher tendency to prefer goal-based action explanations. This work is a necessary step in addressing the challenge of providing personalised explanations in human-robot and human-agent interaction.

Children are eager to anthropomorphize (ascribe human attributes to) social robots. As a consequence they expect a more unconstrained, substantive and useful interaction with the robot than is possible with the current state-of-the art. In this paper we reflect on several of our user studies and investigate the form and role of expectations in child-robot interaction. We have found that the effectiveness of the social assistance of the robot is negatively influenced by misaligned expectations. We propose three strategies that have to be worked out for the management of expectations in child-robot interaction: 1) be aware of and analyze children's expectations, 2) educate children, and 3) acknowledge robots are (perceived as) a new kind of `living' entity besides humans and animals that we need to make responsible for managing expectations.

When an agent program exhibits unexpected behaviour, a developer needs to locate the fault by debugging the agent’s source code. The process of fault localisation requires an understanding of how code relates to the observed agent behaviour. The main aim of this paper is to design a source-level debugger that supports single-step execution of a cognitive agent program. Cognitive agents execute a decision cycle in which they process events and derive a choice of action from their beliefs and goals. Current state-of-the-art debuggers for agent programs provide insight in how agent behaviour originates from this cycle but less so in how it relates to the program code. As relating source code to generated behaviour is an important part of the debugging task, arguably, a developer also needs to be able to suspend an agent program on code locations. We propose a design approach for single-step execution of agent programs that supports both code-based as well as cycle-based suspension of an agent program. This approach results in a concrete stepping diagram ready for implementation and is illustrated by a diagram for both the Goal and Jason agent programming languages, and a corresponding full implementation of a source-level debugger for Goal in the Eclipse development environment. The evaluation that was performed based on this implementation shows that agent programmers prefer a source-level debugger over a purely cycle-based debugger.

We propose a new strategic model of negotiation, called Boolean negotiation games. Our model is inspired by Boolean games and the alternating offers model of bargaining. It offers a computationally grounded model for studying properties of negotiation protocols in a qualitative setting. Boolean negotiation games can yield agreements that are more beneficial than stable solutions (Nash equilibria) of the underlying Boolean game.

In search and rescue missions robots are used to help rescue workers in exploring the disaster site. Our research focuses on how multiple robots and rescuers act as a team, and build up situation awareness. We propose a multi-agent system where each agent supports one member, either human or robot. For representing high-level information about the mission, we design an ontology that serves as a shared knowledge base for the agents. We investigate how to create agent-based ontological reasoning to provide team members with decision support, automate basic monitoring tasks, and realize the display logic that dictates how to display useful information for the rescuers, based on their different roles, tasks and situations.

A variety of agent programming languages have been proposed but only few comparative studies have been performed to evaluate the strengths and weaknesses of these languages. In order to gain a better understanding of features in and their use by programmers of these languages, we perform a study which compares the two languages GOAL and POSH. The study aims at advancing our knowledge of the benefits of using agentoriented languages and at contributing to the evolution of these languages. The main focus of the study is on the usability of both languages and the differences between novice and more advanced programmers that use either language. As POSH requires Java programming experience, we expected novice POSH programmers to perform better on the tasks than novice GOAL programmers whereas we hypothesized this difference would not be observed between more advanced programmers. However, results suggest that there is no significant difference. The study does suggest that general experience and tooling support can make a difference. Analysis of the tasks and the observations made about the use of the languages, moreover, suggests ways to improve the experimental design in such a way that differences in usability of the frameworks could be established.

A negotiation between agents is typically an incomplete information game, where the agents initially do not know their opponent’s preferences or strategy. This poses a challenge, as efficient and effective negotiation requires the bidding agent to take the other’s wishes and future behavior into account when deciding on a proposal. Therefore, in order to reach better and earlier agreements, an agent can apply learning techniques to construct a model of the opponent. There is a mature body of research in negotiation that focuses on modeling the opponent, but there exists no recent survey of commonly used opponent modeling techniques. This work aims to advance and integrate knowledge of the field by providing a comprehensive survey of currently existing opponent models in a bilateral negotiation setting. We discuss all possible ways opponent modeling has been used to benefit agents so far, and we introduce a taxonomy of currently existing opponent models based on their underlying learning techniques. We also present techniques to measure the success of opponent models and provide guidelines for deciding on the appropriate performance measures for every opponent model type in our taxonomy.

A parameterized behavior model was developed for robots to show mood during task execution. In this study, we applied the model to the coverbal gestures of a robotic storyteller. This study investigated whether parameterized mood expression can 1) show mood that is changing over time; 2) reinforce affect communication when other modalities exist; 3) influence the mood induction process of the story; and 4) improve listeners' ratings of the storytelling experience and the robotic storyteller. We modulated the gestures to show either a congruent or an incongruent mood with the story mood. Results show that it is feasible to use parameterized coverbal gestures to express mood evolving over time and that participants can distinguish whether the mood expressed by the gestures is congruent or incongruent with the story mood. In terms of effects on participants we found that mood-modulated gestures (a) influence participants' mood, and (b) influence participants' ratings of the storytelling experience and the robotic storyteller.

To design good cooperative team members in robotics it is important to know what coordination mechanisms are required. Our approach to explore the need for a coordination mechanism is based on a systematic methodology to identify team coordination requirements. We show that a team combined of robots that each individually can solve a task not always is able to guarantee task completion as a team. In these cases some mechanism for coordination is required and we formally identify various problem classes that impose different requirements. We introduce a formal task model and distinguish between no, implicit and explicit coordination mechanisms. This model is used to study which mechanisms guarantee task completion. It allows us to prove some empirical findings reported in the literature such as that a simple foraging task does not require coordination.

The aim of our work is to design bodily mood expressions of humanoid robots for interactive settings that can be recognized by users and have (positive) effects on people who interact with the robots. To this end, we develop a parameterized behavior model for humanoid robots to express mood through body language. Different settings of the parameters, which control the spatial extent and motion dynamics of a behavior, result in different behavior appearances expressing different moods. In this study, we applied the behavior model to the gestures of the imitation game performed by the NAO robot to display either a positive or a negative mood. We address the question whether robot mood displayed simultaneously with the execution of functional behaviors in a task can (a) be recognized by participants and (b) produce contagion effects. Mood contagion is an automatic mechanism that induces a congruent mood state by means of the observation of another person’s emotional expression. In addition, we varied task difficulty to investigate how the task load mediates the effects. Our results show that participants are able to differentiate between positive and negative robot mood and they are able to recognize the behavioral cues (the parameters) we manipulated. Moreover, self-reported mood matches the mood expressed by the robot in the easy task condition. Additional evidence for mood contagion is provided by the fact that we were able to replicate an expected effect of negative mood on task performance: in the negative mood condition participants performed better on difficult tasks than in the positive mood condition, even though participants’ self-reported mood did not match that of the robot.

Given the growing interest in automated negotiation, the search for effective strategies has produced a variety of different negotiation agents. Despite their diversity, there is a common structure to their design. A negotiation agent comprises three key components: the bidding strategy, the opponent model and the acceptance criteria. We show that this three-component view of a negotiating architecture not only provides a useful basis for developing such agents but also provides a useful analytical tool. By combining these components in varying ways, we are able to demonstrate the contribution of each component to the overall negotiation result, and thus determine the key contributing components. Moreover, we study the interaction between components and present detailed interaction effects. Furthermore, we find that the bidding strategy in particular is of critical importance to the negotiator's success and far exceeds the importance of opponent preference modeling techniques. Our results contribute to the shaping of a research agenda for negotiating agent design by providing guidelines on how agent developers can spend their time most effectively.

Every driver behaves differently in traffic. However, when it comes to micro-simulation of drivers with a high level of detail no framework manages to model the complexities of various driving styles as well as scale up to larger simulations. We propose a framework of micro-simulation combined with cognitive agents to facilitate such simulation tasks. Our goal is to (i) model individual drivers, and (ii) use this framework for the purpose of simulating realistic highway traffic with heterogeneous driving styles. The challenge is therefore to create a framework that facilitates such complex modeling and supports large scale simulations. We evaluate the framework from two perspectives. First, the ability to represent, model and simulate dissimilar drivers in addition to study and compare emerging behavior. Second, the scalability of the framework. We report on our experiences with the framework, outline several challenges and identify future areas for development.

We investigate the role of communication in the coordination of cooperative robot teams and its impact on performance in search and retrieval tasks. We first discuss a baseline without communication and analyse various kinds of coordination strategies for exploration and exploitation. We then discuss how the robots construct a shared mental model by communicating beliefs and/or goals with one another, as well as the coordination protocols with regard to subtask allocation and destination selection. Moreover, we also study the influence of various factors on performance including the size of robot teams, the size of the environment that needs to be explored and ordering constraints on the team goal. We use the Blocks World for Teams as an abstract testbed for simulating such tasks, where the team goal of the robots is to search and retrieve a number of target blocks in an initially unknown environment. In our experiments we have studied two main variations: a variant where all blocks to be retrieved have the same color (no ordering constraints on the team goal) and a variant where blocks of various colors need to be retrieved in a particular order (with ordering constraints). Our findings show that communication increases performance but significantly more so for the second variant and that exchanging more messages does not always yield a better team performance.

This paper reports our investigation into the effects of bodily mood expression of a humanoid robot in a scenario close to real life. To this end, we used the NAO robot to perform as a lecturer in a university class. To display either a positive or a negative mood, we modulated 41 co-verbal gestures by adjusting behavior parameters that control spatial extent and motion dynamics, without modifying gesture function. Unique in this study is that (a) the robot gave an actual lecture to real students, (b) the interaction is one-to-many and relatively long (30 min), and (c) mood modulation was applied to a large set of behaviors. The robot presented the same lecture either in a positive or a negative mood to two audiences (between subjects). Although statistical analysis does not show that participants consciously recognized the robot mood, the results do show that participants in the positive mood condition rated their own arousal significantly higher than in the negative condition. Further, video annotation showed increased valence and arousal of the audience in the positive condition. Finally, participants' ratings of the lecturing quality and the gesture quality of the robot are higher in the positive condition, demonstrating the importance of robot mood expression in a one-to-many interaction setting.

The design of automated negotiators has been the focus of abundant research in recent years. However, due to difficulties involved in creating generalized agents that can negotiate in several domains and against human counterparts, many automated negotiators are domain specific and their behavior cannot be generalized for other domains. Some of these difficulties arise from the differences inherent within the domains, the need to understand and learn negotiators’ diverse preferences concerning issues of the domain, and the different strategies negotiators can undertake. In this paper we present a system that enables alleviation of the difficulties in the design process of general automated negotiators termed Genius, a General Environment for Negotiation with Intelligent multi‐purpose Usage Simulation. With the constant introduction of new domains, e‐commerce and other applications, which require automated negotiations, generic automated negotiators encompass many benefits and advantages over agents that are designed for a specific domain. Based on experiments conducted with automated agents designed by human subjects using Genius we provide both quantitative and qualitative results to illustrate its efficacy. Finally, we also analyze a recent automated bilateral negotiators competition that was based on Genius. Our results show the advantages and underlying benefits of using Genius and how it can facilitate the design of general automated negotiators.

In many logic-based BDI agent programming languages, plan selection involves inferencing over some underlying knowledge representation. While context-sensitive plan selection facilitates the development of flexible, declarative programs, the overhead of evaluating repeated queries to the agent's beliefs and goals can result in poor run time performance. In this paper we present an approach to multi-cycle query caching for logic-based BDI agent programming languages. We extend the abstract performance model presented in (Alechina et al. 2012) to quantify the costs and benefits of caching query results over multiple deliberation cycles. We also present results of experiments with prototype implementations of both single- and multi-cycle caching in three logic-based BDI agent platforms, which demonstrate that significant performance improvements are achievable in practice.

Bodily expression of affect is crucial to human robot interaction. Our work aims at designing bodily expression of mood that does not interrupt ongoing functional behaviors. We propose a behavior model containing specific (pose and motion) parameters that characterize the behavior. Parameter modulation provides behavior variations through which affective behavioral cues can be integrated into behaviors. To investigate our model and parameter set, we applied our model to two concrete behaviors (waving and pointing) on a NAO robot, and conducted a user study in which participants (N=24) were asked to design such variations corresponding with positive, neutral, and negative moods. Preliminary results indicated that most parameters varied significantly with the mood variable. The results also suggest that the relative importance may be different between parameters, and parameters are probably interrelated. This paper presents the analysis of these aspects. The results show that the spatial extent parameters (hand-height and amplitude), the head vertical position, and the temporal parameter (motion-speed) are the most important parameters. Moreover, multiple parameters were found to be interrelated. These parameters should be modulated in combination to provide particular affective cues. These results suggest that a designer should focus on the design of the important behavior parameters and utilize the parameter combinations when designing mood expression.

An affordance is a relation between an object, an action, and the effect of that action in a given environmental context. One key benefit of the concept of affordance is that it provides information about the consequence of an action which can be stored and reused in a range of tasks that a robot needs to learn and perform. In this paper, we address the challenge of the on-line learning and use of affordances simultaneously while performing goal-directed tasks. This requires efficient online performance to ensure the robot is able to achieve its goal fast. By providing conceptual knowledge of action possibilities and desired effects, we show that a humanoid robot NAO can learn and use affordances in two different task settings. We demonstrate the effectiveness of this approach by integrating affordances into an Extended Classifier System for learning general rules in a reinforcement learning framework. Our experimental results show significant speedups in learning how a robot solves a given task.

Bodily expression of affect is crucial to human robot interaction. We distinguish between emotion and mood expression, and focus on mood expression. Bodily expression of an emotion is explicit behavior that typically interrupts ongoing functional behavior. Instead, bodily mood expression is integrated with functional behaviors without interrupting them. We propose a parameterized behavior model with specific behavior parameters for bodily mood expression. Robot mood controls pose and motion parameters, while those parameters modulate behavior appearance. We applied the model to two concrete behaviors - waving and pointing - of the NAO robot, and conducted a user study in which participants (N=24) were asked to design the expression of positive, neutral, and negative moods by modulating the parameters of the two behaviors. Results show that participants created different parameter settings corresponding with different moods, and the settings were generally consistent across participants. Various parameter settings were also found to be behavior-invariant. These findings suggest that our model and parameter set are promising for expressing moods in a variety of behaviors.

The past years have seen increasing cooperation between psychology and computer science in the field of computational modeling of emotion. However, to realize its potential, the exchange between the two disciplines, as well as the intradisciplinary coordination, should be further improved. We make three proposals for how this could be achieved. The proposals refer to: 1) systematizing and classifying the assumptions of psychological emotion theories; 2) formalizing emotion theories in implementation-independent formal languages (set theory, agent logics); and 3) modeling emotions using general cognitive architectures (such as Soar and ACT-R), general agent architectures (such as the BDI architecture) or general-purpose affective agent architectures. These proposals share two overarching themes. The first is a proposal for modularization: deconstruct emotion theories into basic assumptions; modularize architectures. The second is a proposal for unification and standardization: Translate different emotion theories into a common informal conceptual system or a formal language, or implement them in a common architecture.

We present a modified version of Extended Classifier System (XCS) on a humanoid NAO robot. The robot is capable of learning a complete, accurate, and maximally general map of an environment through evolutionary search and reinforcement learning. The standard alternation between explore and exploit trials is revised so that the robot relearns only when necessary. This modification makes the learning more effective and provides the XCS with external memory to evaluate the environmental change. Furthermore, it overcomes the drawbacks of learning rate settings in traditional XCS. A simple object seeking task is presented which demonstrates the desirable adaptivity of LCS for a sequential task on a real robot in dynamic environments.

The “negotiation dance”, as Raiffa calls the dynamic pattern of the bidding, has an important influence on the outcome of the negotiation. The current practice of evaluating a negotiation strategy is to focus on fairness and quality aspects of the ag

An important problem in agent verification is a lack of proper understanding of the relation between agent programs on the one hand and agent logics on the other. Understanding this relation would help to establish that an agent programming language is both conceptually well-founded and well-behaved, as well as yield a way to reason about agent programs by means of agent logics. As a step toward bridging this gap, we study several issues that need to be resolved in order to establish a precise mathematical relation between a modal agent logic and an agent programming language specified by means of an operational semantics . In this paper, we present an agent programming theory that provides both an agent programming language as well as a corresponding agent verification logic to verify agent programs. The theory is developed in stages to show, first, how a modal semantics can be grounded in a state-based semantics, and, second, how denotational semantics can be used to define the mathematical relation connecting the logic and agent programming language. Additionally, it is shown how to integrate declarative goals and add precompiled plans to the programming theory. In particular, we discuss the use of the concept of higher-order goals in our theory. Other issues such as a complete axiomatization and the complexity of decision procedures for the verification logic are not the focus of this paper and remain for future investigation.