ABSTRACT. Understanding human motor learning and adaptation processes is an integral step in developing rehabilitative engineering solutions and training strategies for assistive technologies. Natural skill acquisition enables continually precise movements despite inherent noise in motor execution, sensory perception, and dynamic changes in body parameters (growth, age, etc.) and the external environment. As an initial step, motor learning research has aimed to identify the mechanisms of natural human adaptation during the acquisition of motor skills. Results presented here confirm existing literature on motor adaptation using a remote web-based experimental paradigm that could provide a valuable option to conduct additional future work with expanded more diverse subject populations.

ABSTRACT. The ubiquitous availability of sensors and smart devices makes IoT networks more and more complex to manage and control. A natural language interface (NLI) would allow users to interact with the devices via human language by translating the user command into a machine-interpretable meaning representation, often called logical forms.

Despite the rapid development of conversational interfaces in smart home and personal intelligent assistant use cases, there are limited research and applications in industrial sensor and actuator networks, usually referred to as Industrial Internet of Things (IIoT). In this paper, we show an early phase design principle of a semantic representation to express IIoT device interactions and propose a data-focused workflow of IIoT automation system architecture.

ABSTRACT. Due to the introduction of more and more new technologies into the work environment, the role of the worker often changes from executing a task to monitoring an automated system. Monitoring tasks, however, can induce high levels of mental workload and thus lead to a decrease in the worker's performance and consequently to an increase in the risk of making serious errors. A promising solution to this problem is the use of assistance systems (AS). As the worker still remains responsible for an error-free operation of the automated system, the AS should be designed human-centered. Hence, it is crucial, to examine the psychological factors that influence the user's acceptance of the AS in order to ensure a successful long-term human-machine collaboration. A concept for a human-centered AS for a monitoring task is presented. Thereby, different assistance types and strategies as well as methods to determine the right point of time to apply these strategies during the task are looked at. Finally, challenges for the development of such an AS are discussed and possible solutions are proposed.

ABSTRACT. In current production environments, industry 4.0 or smart production plays an essential role. The trend towards more and more individualization and an increasing number of product variants place high demands on people working in manufacturing. Assembly workers must achieve increasingly shorter cycle times and constantly learn how to assemble new products. The support from assistive systems (e.g. digital work instructions, pick-by-light, automatic component recognition for quality control) is necessary in these environments. But many systems and prototypes presented in this field work perfectly in lab environments, however, they may fall short when deployed in a real production environment. We developed a comprehensive and flexible assistance system for manufacturing and identified design and technical challenges, which may only appear when working at the shop floor and. In this paper, we present our experiences and discuss our key challenges and lessons learned.

ABSTRACT. In this work, we analyze the suitability of employing visual feedback for pseudo haptics as a replacement of active haptics in an immersive virtual reality (VR) environment. A controller-free gesture interaction is widely considered to be a natural user interface in VR. As the controller is not employed, the lack of active haptic feedback can often result in a frustrating experience for complex dynamic gestures, e.g., grab, squeeze, clasp etc. These actions are very easy to perform using a specialized devices or controllers with active haptic feedback, e.g., data gloves with force feedback or controllers with analog triggers and vibrations can be utilized for an immediate or continuous feedback. In contrast, these mechanisms are completely missing in a controller-free interaction. We present an on-screen visual mechanism as the pseudo haptic feedback of a dynamic squeeze / grab gesture to replace the active haptic feedback. Our proposed approach allows for the continuous visualization of a squeeze / grab gesture. We implemented an interaction mechanism to test the visualization for these dynamic gestures and compared it with a system with no pseudo haptics. The results from the user study show that an on-screen continuous visualization can be used as pseudo haptics for a dynamic squeeze / grab gesture in an immersive VR environment.

ABSTRACT. This contribution presents sample scenarios for laser safety training using virtual reality. The scenarios illustrate activities involving laser devices used in industry, medicine, didactic laboratories and laser shows. The described applications are incorporated into laser safety training conducted by the Central Institute for Labour Protection – National Research institute (CIOP-PIB).

ABSTRACT. Trust is an important emerging area of study in human-robot cooperation. Many studies have begun to look at the issue of robot (agent) capability as a predictor of human trust in the robot. However, the assumption that agent capability is the sole predictor of human trust could underestimate the complexity of the problem. This study aims to investigate the effects of agent-strategy and agent-capability in a visual search task. Fourteen subjects were recruited to partake in a web-based grid search task. They were each paired with a series of autonomous agents to search an on-screen grid to find a number of outlier objects as quickly as possible. Both the human and agent searched the grid concurrently and the human was able to see the movement of the agent. Each trial, a different autonomous agent with its assigned capability, used one of three search strategies to assist their human counterpart. After each trial, the autonomous agent reported the number of outliers it found, and the human subject was asked to determine the total number of outliers in the area. Some autonomous agents reported only a fraction of the outliers they encountered, thus coding a varying level of agent capability. Human subjects then evaluated statements related to the behavior, reliability, and trust of the agent. The results showed increased measures of trust and reliability with increasing capability. Additionally, the most legible search strategies received the highest average ratings in a measure of familiarity. Remarkably, given no prior information about capabilities or strategies that they would see, subjects were able to determine consistent trustworthiness of the agent. Furthermore, both capability and strategy of the agent had statistically significant effects on the human's trust in the agent.

ABSTRACT. Understanding the intellectual landscape of scientific communities and their collaborations has become an indispensable part of research per se. In this regard, measuring similarities among scientific documents can help researchers to identify groups with similar interests as a basis for strengthening collaboration and university-industry linkages. To this end, we intend to evaluate the performance of hybrid crowd-computing methods in measuring the similarity between document pairs by comparing the results achieved by crowds and artificial intelligence (AI) algorithms. That said, in this paper we designed two types of experiments to illustrate some issues in calculating how similar an automatic solution is to a given ground truth. In the first type of experiments, we created a crowdsourcing campaign consisting of four human intelligence tasks (HITs) in which the participants had to indicate whether or not a set of papers belonged to the same author. The second type involves a set of natural language processing (NLP) processes in which we used the TF-IDF measure and the Bidirectional Encoder Representation from Transformers (BERT) model. The results of the two types of experiments carried out in this study provide preliminary insight into detecting major contributions from human-AI cooperation at similarity calculation in order to achieve better decision support. We believe that in this case decision makers can be better informed about potential collaborators based on content-based insights enhanced by hybrid human-AI mechanisms.

ABSTRACT. Internet of Everything (IoE) is a promising paradigm that integrates the Internet of Things (IoT), Industrial Internet, Internet of People, and many Internet-based paradigms to transform the industry, society, and people lives. It provides seamless integration of intelligent devices - with sensing, identification, processing, communication, and networking capabilities, Big Data – with machine learning, analytics, and distributed computing, and human sensors – with collaboration, intelligent cognition and social networks. IoE brings excellent opportunities to improve changes in society with collaborative intelligent systems. However, these new technologies also bring significant challenges and do not address major existing issues, including interoperability, reliability, and intelligence management. Awareness of these issues is required before IoE applications can be widely deployed. This work introduces an observatory for cataloging IoE applications. Registering these applications will support researchers, developers, and users to exchange more information, design improved IoE applications, facilitating the composition of different enablers (human and non-humans). The main contributions of this work include the proposal of a technology platform, the IoE Database (IoEDB) that enables the distributed and collaborative cataloging of IoE initiatives and provides the evolution of a ‘live’ IoE knowledge-based taxonomy to support interoperability and awareness of IoE enablers.

ABSTRACT. In a complex task environment in which team behavior emerges and evolves, team agility is one of the primary determinants of a team’s success. Agility is considered an emergent phenomenon in which lower-level system elements interact to adapt to the dynamic environment. One of the dimensions of team agility is interactive decision-making. In this study, we conceptually model individual team member’s interactive decision-making process for their taskwork; we observe how much the choices of one team member depend on antecedent decisions and the behavior of the other team members. This also help us understand how team members synchronize during the decision-making process in agile teams, especially when team members team up with a machine. To improve the understanding of interactive decision-making, we also propose two modeling techniques: (1) quantum cognition for the taskwork decision-making processes and (2) nonlinear dynamical systems modeling for teamwork.

ABSTRACT. This work presents “LeARn”, a new network-based collaborative learning environment that employs augmented reality to transform a real-world surface in a virtual lab. The system is a contribution towards replacing a face-to-face learning environment with an augmented collaborative setting. To demonstrate the system, a scenario with a virtual chemistry lab is presented. In this demo, any real-world surface is augmented by virtual lab equipment utilized in a chemistry experiment. The virtual lab is hosted by the instructor, and all the students can join the lab only using their mobile phones or tablets. Each member can interact with the lab equipment, which can be visualized in real-time by the instructor or fellow students. The system allows for real-time communication that fosters a true collaborative environment. The resulting system demonstrates that a complex lab experiment can be performed from a personalized space that can incorporate positive traits of a collaborative environment. The system was deployed and evaluated in an uncontrolled user study, and the results show the effectiveness of an AR-based interactive and collaborative learning environment.

ABSTRACT. Human-Centered Computing and AI are two fields devoted to several cross-intersecting interests in the modern AI design. They consider human factors and the machine learning algorithms to enhance compatibility and reliability for human-robot interaction and cooperation. In this work, we propose a novel design concept for the challenging issues that have raised ethical dilemmas; an augmented ethical causality with successor representation for policy gradient models the Human-Centered AI with environments. The proposed system leverages Human-Centered AI using explainable knowledge to construct ethical causations, and shows it significantly outperformed the statistical approach and baselines alone by further considering meta parametric Human-Centered ethical priorities, when compared to other approaches in the simulated game theory environments of Deep Reinforcement Learning. The experimental results aim to efficiently and effectively access the cause, effect and impact of multi-agent heterogeneity in the DRL environments for natural, general and significant causal learning representations.

ABSTRACT. Explainability in AI (XAI) is being investigated in various AI driven systems like search engines to promote trust and fairness in AI. In the context of search systems one major XAI goal is to explain the notion of similarity in text to a nonexpert user like an avid reader searching for books in a digital library. What is the similarity between book A and B with respect to narrative time? How does the emotion change? This is often an important criteria for searching books. In this work we extend a recent explainable AI driven book search engine [1] with local explanations that visually support users to comprehend the similarity model. Contribution of this work lies in use of topic models and sentiment driven features to show how plot development of a fiction book change over the narrative time. Each book is represented as a feature vector in an interpretable feature space with aspects such as sentence complexity, writing style. In the absence of ground truth similarities and explanations, the system is evaluated on the aspects of reliability of search results and explanations in user studies. Different baselines are created such as a pseudo-random ranker and a bag-of-word model. Baselines are compared with our back-end ranking model, keeping the user interface same for all models. Experimental results on lab based user study with eye tracking mechanism indicate that the explanations help towards making the search system reliable for non expert users.

ABSTRACT. This research examines the relationship between anticipatory pushing of information and trust in human–autonomy teaming in a remotely piloted aircraft system - synthetic task environment. Two participants and one AI teammate emulated by a confederate executed a series of missions under routine and degraded conditions. We addressed the following questions: (1) How do anticipatory pushing of information and trust change from human to human and human to autonomous team members across the two sessions? and (2) How is anticipatory pushing of information associated with the trust placed in a teammate across the two sessions? This study demonstrated two main findings: (1) anticipatory pushing of information and trust differed between human-human and human-AI dyads, and (2) anticipatory pushing of information and trust scores increased among human-human dyads under degraded conditions but decreased in human-AI dyads.

ABSTRACT. As the aviation industry is actively working on adopting AI for air traffic, stakeholders agree on the need for a human-centered approach. However, automation design is often driven by user-centered intentions, while the development is actually technology-centered. This can be attributed to a discrepancy between the system designers' perspective and real-world operational complexity. The same can be observed with AI applications where most design efforts focus on the interface between humans and AI, while the overall system design is built on preconceived assumptions. To understand potential usability issues of AI-driven cockpit assistant systems from the users' perspective, we conducted interviews with four experienced pilots. While our participants did discuss interface issues, they were much more concerned about how autonomous systems could be a burden if the operational complexity exceeds their capabilities. Our results thus point toward an area that is often neglected by designers and could cause unexpected issues in real usage. Besides commonly addressed human-AI interface issues, we believe that more consideration for operational complexities on a system-design level is necessary.

ABSTRACT. As technology continues to advance, humans will need to advance with it. This paper starts out by analyzing how humans adopt technology at a societal level. From there, I discuss some of the methods with which we can increase the ability of technology to be adopted: mainly pertaining in education and increasing usability of technology in a general sense. These methods will help bring us to a future paradigm of technology that I think has great benefits- that of ubiquitous computing. Finally, I go over the ethical implications of technological adoption and address it in a contemporary sense.

ABSTRACT. Trust in autonomous teammates has been shown to be a key factor in human-autonomy team (HAT) performance, and anthropomorphism is a closely related construct that is underexplored in HAT literature. This study investigates whether perceived anthropomorphism can be measured from team communication behaviors in a simulated remotely piloted aircraft system task environment, in which two humans in unique roles were asked to team with a synthetic (i.e., autonomous) pilot agent. We compared verbal and self-reported measures of anthropomorphism with team error handling performance and trust in the synthetic pilot. Results for this study show that trends in verbal anthropomorphism follow the same patterns expected from self-reported measures of anthropomorphism, with respect to fluctuations in trust resulting from autonomy failures.

ABSTRACT. Many taxonomies of levels of automation have been presented in the literature; however, the discrete and ordinal nature of these methods may limit reliable prediction of operator performance. This study defined an “automation rate” to quantify the level of automation in systems. To calculate the automation rate it is necessary to: classify all functions in the automation system according to stages of information processing, calculate the automation rate for each stage, set weights for these automation rates, and finally obtain the overall automation rate for the system. The practicality and feasibility of this model are verified through a case study analysis. In addition, this paper proposes a new relationship between the automation rate and operator situation awareness response, based on existing empirical research findings. Through case analysis and mathematical proof, the rationality of the form is demonstrated. This work lays the foundation for subsequent operator performance optimization analysis.