Organizers: Manolis Chiou, Maria Kyrarini, Nicholas Conlon, Andreas Theodorou, Balint Varga, Ayse Kucukyilmaz

Time: August 26th, 2024 full day

Homepage: https://sites.google.com/view/vat2024

As robots are introduced to various domains and applications, Human-Robot Teaming (HRT) capabilities are essential. Such capabilities involve teaming with humans in\on\out-the-loop at different levels of abstraction, leveraging the complementing capabilities of humans and robots. This requires robotic systems with the ability to dynamically vary their level or degree of autonomy to collaborate with the human(s) efficiently and overcome various challenging circumstances. Variable Autonomy (VA) is an umbrella term encompassing such research, including but not limited to shared control and shared autonomy, mixed-initiative, adjustable autonomy, and sliding autonomy.

The workshop is driven by the timely need to bring together VA-related research and practices often disconnected across different communities, as the field is relatively young.

The workshop aims to consolidate research in VA for HRT towards real-world applications. VA offers a promising approach to bridging the gap between teleoperation (the current paradigm in many real-world applications) and full autonomy (which cannot always be realistically deployed and stakeholders often remain sceptical of). To this end, and given the complexity and span of Human-Robot systems, this workshop will adopt a holistic trans-disciplinary approach aiming to a) identify application domains for VA; b) identify and classify related common challenges, opportunities, and the disciplines that need to come together to tackle the challenges; c) define short- and long-term research goals for the community.

To achieve these objectives, this workshop aims to bring together industry stakeholders, researchers from fields under the banner of VA, and specialists from other highly related fields such as human factors and psychology.

Organizers: Micol Spitale, Alyssa Kubota, Patrícia Alves-Oliveira, Hatice Gunes

Home page: https://sites.google.com/view/hri4wellbeing2024/home

Time: August 26th full day

In the last year, the use of Large Language Models (LLMs) has spread rapidly in our society, from information retrieval to text classification, as well as robotic applications. Given these rapid advancements, recent studies have attempted to identify challenges and opportunities for using LLMs in various contexts, such as in healthcare and medicine, among others. These works have highlighted that the use of LLMs in healthcare and well-being pose risks in terms of data privacy, questionable credibility and accuracy of information, data bias, interpretability of LLMs, role of LLMs, and deployment of LLMs. However, there is a lack of analysis and understanding of the challenges and opportunities when using LLMs for robotic applications.

While robots for well-being are becoming an increasingly relevant line of HRI research, as people have shown increased interest in using digital tools to improve their well-being, the use of LLMs in such applications is still unexplored. Therefore, there is a need for discussing and identifying the risks and potentials of using LLMs in human-robot interaction for well-being.

Exploring the challenges and opportunities of using LLMs in robots for well-being is extremely important to gain insights into how these models affect the users’ perception of robots in well-being applications. The conversations held in this workshop could serve as the first stepping stone to guide the HRI community in designing robotic applications for well-being that leverage the use of LLMs in a responsible and ethical manner.

Organizers: Alva Jamina Ka Markelius*, Laetitia Tanqueray, Robert Lowe, Yoon Kyung Lee, Yoonwon Jung, Stefan Larsson

Homepage: https://sites.google.com/view/llmra/start

Time: August 26th full day

The integration of Large Language Models (LLMs) into social robotics relates to a larger trend in generative AI that potentially represents a significant shift in the field of human-robot interaction (HRI) and social robotics. We are currently seeing a large shift in the field of HRI where LLMs are being increasingly implemented in a multitude of different applications for social robotics, ranging from well-being and education to healthcare and business. This research trend offers several notable advantages, such as personalising and improving interactions, in particular open-domain dialogue, improving sentiment analysis for affective and emotional interaction as well as reducing reliance on manual methods such as the Wizard of Oz technique. However, integrating LLMs with social robots brings new complexities in design, particularly in shaping robot personality and behaviour to align with the diverse social and cultural norms and contexts. This trend carries ethical and societal concerns and it is vital to rigorously examine the potential risks and ethical issues associated with using LLMs in social robots, especially considering the impact on vulnerable communities.

The objectives of this workshop are to engage with questions arising at the intersection of LLMs and HRI, and in particular related to social impact, ethical considerations and design of personality and behaviour. Our workshop organisers as well as tentatively confirmed speakers represent multiple institutions, organisations and nations such as OpenAI, Seoul National University, Tokyo Tech, Cambridge and KTH. Thus, we aim to create a platform to bring rich, diverse and interdisciplinary perspectives and audiences from around the world to participate in exploring the topic from a human centered perspective.

Organizers: Francesca Fracasso, Alessandro Umbrico, Laura Fiorini, Alessandra Sorrentino

Homepage: https://altruist21.istc.cnr.it

Time: August 26th full day

Demographic change and epidemiological challenges confront society with the need to provide high-quality and sustainable healthcare over time. Shortages of medical personnel, rising healthcare costs and the growth of vulnerable groups are driving the growing adoption of social robots in the healthcare sector. These innovative digital and robotic systems aim to improve care by supporting both staff and patients, tackling tasks ranging from health education to emotional care and medication administration.

However, the interaction between humans and robots in healthcare presents several challenges, including safety, acceptance, and concerns about the replacement of human personnel. Involving users from the beginning is crucial to foster acceptance of these technologies. Furthermore, there is a need to develop innovative approaches, including self-management, to respond to the growing needs for personalized care. Effectively addressing the growing demand for care requires a multidisciplinary approach that combines artificial intelligence, IoT, robotics and social sciences. Only through this synergy it will be possible to offer innovative and impactful solutions to face the healthcare challenges of the future.

Organizers: Jauwairia Nasir, Barbara Bruno, Utku Norman, Muneeb Ahmad, Hifza Javed, Elisabeth Andre

Homepage: https://sites.google.com/view/tsar-workshop

Time: August 26th Morning session

One of the growing trends in social robotics is their use as socially assistive agents, i.e. agents assisting via social interaction. Application areas include the after-care management of patients coming out of neurological disorders such as depression, facilitating therapists in psychotherapy when dealing with children with special education needs and psychopathology; the assistance to elderly navigating through loneliness, cognitive decline, or physical disability. When considering human health professionals, regardless of the assistive goal pursued, real-time awareness of the patient’s physical and psychological status always play a key role in the caregiver’s decision-making process.

From a psychological perspective, this observation poses a number of questions. To what extent artificial agents should and can imitate human caregivers? Would people want a machine to know/judge their psychological status? Would they trust the machine’s judgment? How would users, both caregivers and care recipients, envision and wish interactions with such agents to be?From a technical perspective, this observation raises the need to (i) develop methods for the perception and modeling of relevant human behaviors/states, (ii) devise effective decision-making strategies, and (iii) design user-centered agent behaviors. Furthermore, with the rise of transformative technology such as large language models (LLMs) and their rapid integration in the HRI community, it is inevitable to discuss the impact that they can have on the domain under discussion, particularly because a more vulnerable population is involved.

This workshop aims to bring together the two aforesaid perspectives to advance research in user-centered socially assistive agents and foster this community’s interdisciplin

Organizers: Alessandra Rossi, Patrick Holthaus, Gabriella Lakatos, Sílvia Moros

Homepage: https://scrita.herts.ac.uk

Time: August 26th Afternoon Session

State-of-the-art literature agrees that people’s ability to accept and trust robots is a fundamental prerequisite for a fruitful and successful coexistence between humans and robots. The previous editions of this workshop have highlighted how huge advances have been made in studying and evaluating the factors affecting people’s acceptance and trust in robots in controlled or short-term (repeated interactions) settings. At the same time, they also agree that several open challenges for scientists in robotics, AI and HRI need to be overcome to develop service, personal and collaborative robots that can calibrate and maintain people’s acceptance and trust in robots. In particular, these robots need to be able to proactively adapt their behaviours to the situational context, and people’s social and tasks-related expectations. Another very important aspect is that the field of HRI still lacks metrics that allow for an effective and unmistakable assessment of people’s trust towards robots. During last year's workshop edition (i.e., SCRITA@RO-MAN 2023), together with leading researchers and exceptional speakers from various fields, we started working towards developing such novel methods. We outlined current methods and their strengths, discussing how these measures do not always reflect appropriately, or how some questions might be ambiguous and leave room for interpretation by individual participants. We identified five main factors affecting trust to be investigated to generate a new metric that allows researchers to assess and reduce common side effects influencing how people put their trust in robots.

Organizers: Francesco Vigni, Antonio Andriella, Andrea Rezzani, Jauwairia Nasir, Alyssa Kubota, Silvia Rossi

Time: August 26th Afternoon Session

Homepage: https://warn-ws.github.io/

The importance of personalisation in Human-Robot Interaction has already shown its advantages in multiple scenarios and will become a prevalent direction for the field. Robots are required to adapt their behaviour in both short- and long-term interactions. In the short term, as the interactions are very often limited in time, robots need to learn from scratch the user's preferences and adapt quickly to them. In the long term, users' needs may change and robots need to continuously adapt in a way that keeps them engaged and interested over time. Personalisation can greatly improve short- and long-term interactions in various real-world scenarios by fostering trust and rapport, increasing adherence to the interaction, enhancing engagement through tailored content, and improving task performance. Nonetheless, it is essential to consider whether and to what extent personalisation can be beneficial for interactions and users. Robots developed as end-to-end systems for conducting social interactions can amplify cultural biases, gender and age stereotypes. Therefore, it is crucial to discuss when personalisation is desired or required, and when it should be avoided. In contexts such as healthcare and education, personalisation can lead to inadequate care or support and lower acceptance of the professionals who use the technology (teachers and healthcare professionals). Additionally, collecting personal data to provide tailored assistance can raise privacy concerns, as many machine learning algorithms are not transparent to users. Furthermore, deep learning algorithms may amplify existing biases, hindering the primary goal of making interactions more engaging and trustworthy.

Organizers: Yuchong Zhang. Elmira Yadollahi, Yong Ma, Di Fu, Iolanda Leite, Danica Kragic

Homepage: https://sites.google.com/view/interaiworkshops

Time: August 26th, Morning Session

This workshop aims to explore and discuss the advancements and challenges in human-centered interactive artificial intelligence (AI) within the field of human-robot interaction (HRI). It will focus on the integration of AI technologies that enhance human-robot collaboration, ensuring these interactions are intuitive, efficient, and tailored to human needs and behaviors.

Based on literature, human-centered interactive AI is defined as an AI that enables interactive exploration and manipulation in real-time and is designed with a clear purpose for human benefit while being transparent about who has control over data and algorithms.

This workshop is dedicated to exploring the cutting-edge developments in AI that prioritize interactive, real-time exploration and manipulation, all within the sphere of HRI. We aim to address the design considerations that make AI systems transparent, particularly in terms of data control and algorithmic operations, ensuring that users understand and trust the technology they interact with in the context of human-centered robotics.

As it’s the first edition, this workshop will serve as a platform for fostering innovation, collaboration, and discussion among the HRI community, driving forward the development of human-centered interactive AI in robotics.

Through a series of keynotes, paper presentations, panel discussions, and interactive sessions, we aspire to foster a deep understanding of how human-centered interactive AI can be effectively integrated into HRI systems to create more effective, ethical, and user-friendly interactions.

Organizers: Claire Yilan Liang, Franziska Babel, Hannah Pelikan, Zhi Tan

Homepage: https://sites.google.com/cornell.edu/hriinpublic/home

Time: August 26th, Morning Session

Both research and commercial robots are increasingly deployed in the wild for tasks such as providing information, transporting goods or performing cleaning tasks. While these robots are typically developed for specific tasks and intended users, in the wild, they inevitably encounter a broad variety of people who may behave in ways that are unexpected by robots and roboticists. As these unexpected events are typically anecdotal and typically related to failure, they are often treated as outliers, omitted from quantitative datasets, and thus lack the opportunity for broader discussion.

Although these events are infrequent, researchers who are experienced in deploying and studying robots outside the lab may see recurring patterns and be able to formulate lessons learned. The goal of this workshop is to create a venue for researchers working on `in the wild HRI' to share these unexpected events, identify potential new research directions, and formulate best practices. Additionally, we intend to use this opportunity to create a community-wide means to collect and make documentation of these unexpected behaviors available to all HRI researchers. With this widely-available resource, the events can then be categorized (e.g., underlying causes, external influences). The meta-analyses of this data can be used to uncover themes within these unexpected events and ultimately guide design towards socially fluent robots.

Organizers: Mariacarla Staffa, Silvia Rossi, Alessandra Sciutti

Homepage: https://sites.google.com/view/bailar2024/home

Time: August 26th Morning session

The BAILAR Workshop aims to spotlight the pivotal role of the User in Human-Robot Interaction (HRI) within Assistive Scenarios, where adaptation and learning are essential functionalities. Central to this discussion are nonverbal communication's high-level abilities, particularly focusing on Theory of Mind and the mutual understanding of affective/emotional states between robots and humans. The workshop will delve into methodologies and technologies utilized for detecting and adapting to users' mental states, emotions, and dispositions during HRI, fostering dialogue on experimental protocols and results. Additionally, there will be exploration into potential effects of gender, age, personality, and pathology on the perception of robots from an emotional and affective standpoint. Ethical considerations surrounding the learning and utilization of personal data within assistive applications will also be addressed. The workshop serves as a collaborative platform for a diverse audience, including roboticists, psychologists, computer scientists, social scientists, and ethicists. The discussions will revolve around the roles of adaptation, learning, emotional communication, and mutual affective understanding in HRI for real-world assistive applications. Special attention will be paid to their potential impact on users' acceptance, focusing on aspects such as usability, efficiency, empathy, and emotional responses.