Umeå University's logo

In recent years, there has been growing interest in developing robots capable of explaining their behavior, thereby improving their acceptance by humans with whom they share their environment. Proposed software designs are typically based on the advances being made in conversational systems built on deep learning techniques. However, apart from the ability to formulate explanations, the robot also needs an internal episodic memory, where it stores information from the continuous stream of experiences. Most previous proposals are designed to deal with short streams of episodic data (several minutes long). With the aim of managing larger experiences, we propose in this work a high-level episodic memory, where relevant events are abstracted to natural language concepts. The proposed framework is intimately linked to a software architecture in which the explanations, whether externalized or not, are shaped internally in a collaborative process involving the task-oriented software agents that make up the architecture. The core of this process is a runtime knowledge model, employed as working memory whose evolution allows for capturing the causal events stored in the episodic memory. We present several use cases that illustrate how the suggested framework allows an autonomous robot to generate correct and relevant explanations of its actions and behavior.

This study investigates user experiences of interactions with two types of robots: Pepper, a social humanoid robot, and Double 3, a self-driving telepresence robot. Conducted in a controlled setting with a specific participant group, this research aims to understand how the design and functionality of these robots influence user perception, interaction patterns, and emotional responses. The findings reveal diverse participant reactions, highlighting the importance of adaptability, effective communication, autonomy, and perceived credibility in robot design. Participants showed mixed responses to human-like emotional displays and expressed a desire for robots capable of more nuanced and reliable behaviors. Trust in robots was influenced by their perceived functionality and reliability. Despite limitations in sample size, the study provides insights into the ethical and social considerations of integrating AI in public and professional spaces, offering guidance for enhancing user-centered designs and expanding applications for social and telepresence robots in society.

The deployment of robots in environments shared with humans implies that they must be able to justify or explain their behavior to nonexpert users when the user, or the situation itself, requires it. We propose a framework for robots to generate personalized explanations of their behavior by integrating cause-and-effect structures, social roles, and natural language queries. Robot events are stored as cause–effect pairs in a causal log. Given a human natural language query, the system uses machine learning to identify the matching cause-and-effect entry in the causal log and determine the social role of the inquirer. An initial explanation is generated and is then further refined by a large language model (LLM) to produce linguistically diverse responses tailored to the social role and the query. This approach maintains causal and factual accuracy while providing language variation in the generated explanations. Qualitative and quantitative experiments show that combining the causal information with the social role and the query when generating the explanations yields the most appreciated explanations.

The Umeå Student Conference in Computing Science (USCCS) is organized annually as part of a course given by the Computing Science department at Umeå University. The objective of the course is to give the students a practical introduction to independent research, scientific writing, and oral presentation.

A student who participates in the course first selects a topic and a research question that he or she is interested in. If the topic is accepted, the student outlines a paper and composes an annotated bibliography to give a survey of the research topic. The main work consists of conducting the actual research that answers the question asked, and convincingly and clearly reporting the results in a scientific paper. Another major part of the course is multiple internal peer review meetings in which groups of students read each others’ papers and give feedback to the author. This process gives valuable training in both giving and receiving criticism in a constructive manner. Altogether, the students learn to formulate and develop their own ideas in a scientific manner, in a process involving internal peer reviewing of each other’s work and under supervision of the teachers, and incremental development and refinement of a scientific paper.

Each scientific paper is submitted to USCCS through an on-line submission system, and receives reviews written by members of the Computing Science department. Based on the review, the editors of the conference proceedings (the teachers of the course) issue a decision of preliminary acceptance of the paper to each author. If, after final revision, a paper is accepted, the student is given the opportunity to present the work at the conference. The review process and the conference format aims at mimicking realistic settings for publishing and participation at scientific conferences.

USCCS is the highlight of the course, and this year the conference received 5 submissions (out of a possible 9), which were carefully reviewed by the reviewers listed on the following page.

We are very grateful to the reviewers who did an excellent job despite the very tight time frame and busy schedule. As a result of the reviewing process, 4 submissions were accepted for presentation at the conference. We would like to thank and congratulate all authors for their hard work and excellent final results that are presented during the conference.

This paper presents a taxonomy of agents’ embodiment in physical and virtual environments. It categorizes embodiment based on five entities: the agent being embodied, the possible mediator of the embodiment, the environment in which sensing and acting take place, the degree of body, and the intertwining of body, mind, and environment. The taxonomy is applied to a wide range of embodiment of humans, artifacts, and programs, including recent technological and scientific innovations related to virtual reality, augmented reality, telepresence, the metaverse, digital twins, and large language models. The presented taxonomy is a powerful tool to analyze, clarify, and compare complex cases of embodiment. For example, it makes the choice between a dualistic and non-dualistic perspective of an agent’s embodiment explicit and clear. The taxonomy also aided us to formulate the term “embodiment by proxy” to denote how seemingly non-embodied agents may affect the world by using humans as “extended arms”. We also introduce the concept “off-line embodiment” to describe large language models’ ability to create an illusion of human perception.

The latest developments of chatbots driven by Large Language Models (LLMs), more specifically ChatGPT, have shaken the foundations of how text is created, and may drastically reduce and change the need, ability, and valuation of human writing. Furthermore, our trust in the written word is likely to decrease, as an increasing proportion of all written text will be AI-generated – and potentially incorrect. In this essay, I discuss these implications and possible scenarios for us humans, and for AI itself.

The recent development of large language models (LLMs) and improvements in speech recognition have made it realistic to envision AI-driven robots replace humans in debates, or even debating with each other. One application area is politics, where debating robots could support human politicians and, in principle, they could also debate on their own. However, political debating are highly complex and is often guided by values and ideologies, rather than rational decisions based on explicit facts.

In this paper we discuss how introducing appropriate bias into an LLM can be a way to accomplish this. As a simple proof of concept, we present a novel system of three robots conducting verbal debates on selectable topics. The robots are driven by the LLM GPT-3.5, and the desired political view, level of knowledge, and speaking style of each robot are configurable. The results demonstrate how LLMs may be used to argue both for and against different standpoints in debates, and how the output arguments depend on a programmed bias reflecting desired values and ideological principles.

One important aspect when designing understandable robots is how robots should communicate with a human user to be understood in the best way. In elder care applications this is particularly important, and also difficult since many older adults suffer from various kinds of impairments. In this paper we present a solution where communication modality and communication parameters are adapted to fit both a user profile and an environment model comprising information about light and sound conditions that may affect communication. The Rasa dialogue manager is complemented with necessary functionality, and the operation is verified with a Pepper robot interacting with several personas with impaired vision, hearing, and cognition. Several relevant ethical questions are identified and briefly discussed, as a contribution to the WARN workshop.

In this paper we propose a method to augment a Reinforcement Learning agent with legibility. This method is inspired by the literature in Explainable Planning and allows to regularize the agent’s policy after training, and without requiring to modify its learning algorithm. This is achieved by evaluating how the agent’s optimal policy may produce observations that would make an observer model to infer a wrong policy. In our formulation, the decision boundary introduced by legibility impacts the states in which the agent’s policy returns an action that is non-legible because having high likelihood also in other policies. In these cases, a trade-off between such action, and legible/sub-optimal action is made. We tested our method in a grid-world environment highlighting how legibility impacts the agent’s optimal policy, and gathered both quantitative and qualitative results. In addition, we discuss how the proposed regularization generalizes over methods functioning with goal-driven policies, because applicable to general policies of which goal-driven policies are a special case.