Autonomous robot operations in the industry are becoming increasingly complex. It is therefore a significant challenge to comprehend the fundamental processes and to gain an understanding of the status of these systems. The RoboCup Logistics League (RCLL) represents a small smart factory environment with several workstations and operating robots. Despite its small scale the processes that occur within the league are very complex. Even with live commentary, observers have difficulties to follow the processes and game progress. This results in low interest in the RCLL and only few of visitors at competitions. To address this, we want to present RCLL-AR, an augmented reality (AR) solution visualizing highly relevant information of the RoboCup Logistics League. By using RCLL-AR, spectators of the game can see the current progress of the game, receive additional information about different workstations and understand future robot movements. To gain insights into the benefits of RCLL-AR for different stakeholders, we conducted expert interviews, a novice user study and an HMD study. Our findings showcase challenges AR faces in complex autonomous systems but also indicate benefits for novices and experts.

Robots deployed in human spaces are required to achieve their goals safely and efficiently. Those two objectives can interfere with each other, as safety often requires slowing down, stopping, or taking detours. This is complicated even more by the fact that safety is not only a physical consideration, but also a psychological one: even if a robot is able to stop reliably before any collision, driving at high speeds towards humans will lead to low perceived safety, reducing comfort and increasing the risk of unpredictable reactions to the robot. To tackle this problem, we propose to build on an existing technique for safe path following in dynamic environment, and extend it with safety constraints suitable for social robotics environment. We also focus on making those constraints easily configurable to adapt to different use-cases. We evaluate our proposed method against baselines on simulated environments, integrating established metrics for performance, safety, and comfort.

This paper presents the development of an innovative communication interface between humans and robots, designed for human-in-the-loop interaction with a user interface through natural language. The novelty of the presented approach lies in integration of a Belief-Desire-Intention agent, communicating directly to a robot and ensuring safety properties and verifiable decision making, with large language models that excel in language understanding and generation. We establish a framework that leverages the strengths of both paradigms by allowing users to formulate commands in natural language, using the ability of large language models to interpret and ground them into actionable goals, with the proven ability of Belief-Desire Intention agents to perform verifiable reasoning and goal management. In addition, we utilize the ability of this agent to represent and store its mind state, including its belief base, plan library, and history of selected events and actions, to allow large language models based explanations of the system behavior. Our findings demonstrate that this architecture provides several benefits in terms of performance and safety, without impacting efficiency or requiring extensive integration effort. This research contributes a novel perspective on the combination of large language models with symbolic approaches for explainable Artificial Intelligence and aims at inspiring new developments in human-centered Artificial Intelligence systems.

This paper addresses the capability of autonomous robots to achieve flexible goals in dynamic environments. In such a setting numerous challenges jeopardize the robustness of such systems. Thus, we propose a hierarchical diagnosis concept for layered control architectures, that can detect and deal with such challenges to maintain a consistent knowledge about the world and to allow reliable decision-making. Layered control systems use various knowledge representations and decision-making mechanisms teamed with specialized isolated fault-handling approaches. However, some issues can only be identified if the information from different layers is combined. Our approach addresses challenges like failing actions, uncertain observations, and unmodeled events by propagating observations and diagnoses results throughout the hierarchy. This enhances adaptability and dependability in various domains. In this paper, we present a prototype architecture following this approach.

Efficient and reliable navigation in off-road environments poses a significant challenge for robotics, especially when factoring in the varying capabilities of robots across different terrains. To achieve this, the robot system’s traversability is usually estimated to plan traversable routes through an environment. This paper presents a new approach that utilizes Deep Multimodal Variational Autoencoders (DMVAEs) for estimating the traversability of different robots in complex offroad terrains. Our method utilizes DMVAEs to capture essential environmental information and robot properties, effectively modeling factors that influence robotic traversability. The key contribution of this research is a two-stage traversability estimation framework for various robots in diverse off-road conditions that integrates robot properties in addition to environmental information to predict the traversability for various robots in a single model. We validate our method through real-world experiments involving four ground robots navigating an alpine environment. Comparative evaluations against state-of-the-art traversability estimation methods demonstrate the superior accuracy and robustness of our approach. Additionally, we investigate the transfer of trained models to new robots, enhancing their traversability estimation and extending the applicability of our framework.

The field of robotics has shown significant advances in autonomous systems, particularly in robot navigation. Since the decisions made during navigation can be difficult for human operators to understand, research aims to provide explanations that improve human-robot interaction (HRI). However, generating and designing such explanations with the intention of improving robot-related human factors is still an ongoing research challenge. This paper addresses this challenge by investigating the impact of different explanation types on a set of human factors in the context of robot navigation. For this purpose, we conducted a user study that examined the impact of six different explanation types on commonly used human factors, including trust, satisfaction, situation awareness, likeability, understandability, and perceived usefulness. Additionally, the study provides indications of their general applicability for robot navigation explanations through creation of sum ranks across the observed human factor metrics. The results show that depending on the chosen explanation type, a significant impact on the measured factors can be observed. While constraint-based explanations are generally rated highly across all factors, apologetic explanations are not perceived well across all measured human factors. Our results provide insights into the impact of explanation types used for robot navigation scenarios on robot-related human factors, and also provide practical insights for designing explanations for robot navigation scenarios.

In the field of off-road navigation, where traditional maps often fall short, intuitive and efficient path planning is essential for autonomous off-road vehicles. Navigating in off-road terrain poses unique challenges, requiring innovative solutions for users to understand and trust path suggestions made by an autonomous system. In this paper, we explore the integration of Explainable AI into off-road navigation systems to better understand the complexity of off-road environments. Our research introduces a method tailored to generate contextual explanations for chosen paths using terrain features, environmental factors, and robot capabilities. By combining inverse optimization techniques with shortest path algorithms, our approach aims to answer the question "Why is path p*recommended over path p′, which was expected by the user?" These explanations aim to shed light on the process of a robot’s path planning task, focusing on elevation changes, terrain obstacles, and optimal path choices, thus improving the user’s understanding of the chosen paths. A short user study evaluating the provided explanations generated in different off-road environments validates the effectiveness of our explanation algorithm and shows that it contributes to understanding the planning process of off-road navigation systems.

There is an interest in and need for the use of semiautonomous robots in various fields, such as disaster response. Over the years, different techniques for semiautonomous control were developed, ranging from teleoperation guidance to different user interface designs for setting task constraints and goals interactively. Among those, affordance templates emerged as a recent method for users to efficiently provide robots with contextual information about object shapes, properties, and affordances. In many fields where direct teleoperation is common, affordance templates seem to be a promising candidate for improving performance and usability. However, despite the reports on the potential benefits of this technique in comparison to direct teleoperation, they are often qualitative or focus on tasks where teleoperation is particularly challenging. This can be a problem because task difficulty can influence different performance metrics and human factors, so results from studies that show large differences in task difficulty between interaction modes cannot be directly generalized to tasks with smaller or non-existent differences in difficulty. In this study, we aim to evaluate the effectiveness of affordance templates in the context of debris removal for disaster response, a task where direct teleoperation is a viable technique. We compared the two methods in a simulated setting through a user study involving 41 participants by measuring a) usability through questionnaires and b) performance on secondary tasks, an established measure of spare information processing capacity. The study results show that despite similar difficulty, users performed better on secondary tasks when using affordance template-based semiautonomy in this setting.

The outdoor navigation capabilities of ground robots have improved significantly in recent years, opening up new potential applications in a variety of settings. Cost-based representations of the environment are frequently used in the path planning domain to obtain an optimized path based on various objectives, such as traversal time or energy consumption. However, obtaining such cost representations is still cumbersome, particularly in outdoor settings with diverse terrain types and slope angles. In this paper, we address this problem by using a data-driven approach to develop a cost representation for various outdoor terrain types that supports two optimization objectives, namely energy consumption and traversal time. We train a supervised machine learning model whose inputs consists of extracted environment data along a path and whose outputs are the predicted energy consumption and traversal time. The model is based on a ResNet neural network architecture and trained using field-recorded data. The error of the proposed method on different types of terrain is within 11% of the ground truth data. To show that it performs and generalizes better than currently existing approaches on various types of terrain, a comparison to a baseline method is made.

Abstract Robot localization is a fundamental capability of all mobile robots. Because of uncertainties in acting and sensing, and environmental factors such as people flocking around robots, there is always the risk that a robot loses its localization. Very often behaviors of robots rely on a reliable position estimation. Thus, for dependability of robot systems it is of great interest for the system to know the state of its localization component. In this paper we present an approach that allows a robot to asses if the localization is still correct. The approach assumes that the underlying localization approach is based on a particle filter. We use deep learning to identify temporal patterns in the particles in the case of losing/lost localization. These patterns are then combined with weak classifiers from the particle set and sensor perception for boosted learning of a localization estimator. Through the extraction of features generated by neural networks and its usage for training strong classifiers, the robots localization accuracy can be estimated. The approach is evaluated in a simulated transport robot environment where a degraded localization is provoked by disturbances cased by dynamic obstacles. Results show that it is possible to monitor the robots localization accuracy using convolutional as well as recurrent neural networks. The additional boosting using Adaboost also yields an increase in training accuracy. Thus, this paper directly contributes to the verification of localization performance.

This Innovative Practice Work in Progress paper presents a novel educational project, aiming at the development and implementation of a professional, standardized, internationally accepted system for training and certifying teachers, school students and young people in Artificial Intelligence (AI) and Robotics. In recent years, AI and Robotics have become major topics with a huge impact not only on our everyday life but also on the working environment. Hence, sound knowledge about principles and concepts of AI and Robotics are key skills for the 21st century. Nonetheless, hardly any systematic approaches exist that focus on teaching AI/Robotics principles at K-12 level, addressing both teachers and students. In order to meet this challenge, the European Driving License for Robots and Intelligent Systems is under development. It is based on a number of previously implemented and evaluated projects and comprises teaching curricula and training modules for AI/Robotics, following a competency based, blended learning approach. Additionally, a certification system proves peoples' competencies acquired during the training. By applying this innovative approach - a standardized and widely recognized training and certification system for AI and Robotics at K-12 level for both teachers and students - we envision to foster AI/Robotics literacy on a broad basis.

This Research to Practice Work in Progress paper presents an innovative, scientifically evaluated Educational Robotics project aiming at teaching fundamental Robotics topics to gifted high school students. Robotics plays an increasingly important role in our life. The increased need for well-trained people underlines the importance of a systematic education in Robotics, from K-12 to university. In recent years, a number of Educational Robotics projects and initiatives have emerged, mainly focusing on introducing children and youth to Robotics, but without going into greater detail on fundamental Robotics concepts. In order to close the gap between these introductory Robotics courses for beginners at K-12 level and common Robotics university courses, the MINT-Robo project was developed. Based on previously implemented K-12 Robotics projects, it is designed around the central topic of autonomous mobile robots. Complex contents, usually taught at university level, are prepared for the project's target group. Following the principles of constructionism, the project comprises a variety of hands-on elements, applying a project-based learning approach. The first pilot course was implemented and evaluated in form of youth research week in summer 2018. First results indicate that the goal of imparting fundamental Robotics concepts to talented high school students was achieved.

Artificial Intelligence (AI) is becoming increasingly important. Thus, sound knowledge about the principles of AI will be a crucial factor for future careers of young people as well as for the development of novel, innovative products. Addressing this challenge, we present an ambitious 3-year project focusing on developing and implementing a professional, internationally accepted, standardized training and certification system for AI which will also be recognized by the industry and educational institutions. The approach is based on already implemented and evaluated pilot projects in the area of AI education. The project’s main goal is to train and certify teachers and mentors as well as students and young people in basic and advanced AI topics, fostering AI literacy among this target audience.

Complex technical systems usually show a dynamic behavior that is often conveniently represented with a discrete event model. Such a behavior is the result of dynamic components which interact with each other. Due to the complexity of technical systems faults are not totally avoidable. In order to deal with such faults diagnosing the system at run-time is of great interest. To perform such a diagnosis it is common to use fault models. Such models are in practice often hard to obtain. To address this problem we show a diagnosis approach for discrete event systems which uses the model of the nominal behavior only. In order to perform this diagnosis we adopt the well known idea of consistency based diagnosis.

Profound knowledge about Artificial Intelligence (AI) will become increasingly important for careers in science and engineering. Therefore an innovative educational project teaching fundamental concepts of AI at high school level will be presented in this paper. We developed an AI-course covering major topics (problem solving, search, planning, graphs, datastructures, automata, agent systems, machine learning) which comprises both theoretical and hands-on components. A pilot project was conducted and empirically evaluated. Results of the evaluation show that the participating pupils have become familiar with those concepts and the various topics addressed. Results and lessons learned from this project form the basis for further projects in different schools which intend to integrate AI in future secondary science education.

Artificial Intelligence (AI) already plays a major role in our daily life (e.g. intelligent household appliances like robotic vacuum cleaners or AI-based applications like Google Maps, Google Now, Siri, Cortana, ...). Sound knowledge about AI and the principles of computer science will be of vast importance for future careers in science and engineering. Looking towards the near future, jobs will largely be related to AI. In this context literacy in AI and computer science will become as important as classic literacy (reading/writing). By using an analogy with this process we developed a novel AI education concept aiming at fostering AI literacy. The concept comprises modules for different age groups on different educational levels. Fundamental AI/computer science topics addressed in each module are, amongst others, problem solving by search, sorting, graphs and data structures. We developed, conducted and evaluated four proof-of-concepts modules focusing on kindergarten/primary school as well as middle school, high school and university. Preliminary results of the pilot implementations indicate that the proposed AI education concept aiming at fostering AI literacy works.

When autonomous robots are deployed in an industrial setting they are expected to work 24 hours a day, 7 days a week. Therefore, dependability of the robots is crucial. In this paper we present an approach following the model-driven engineering idea that supports dependability in different stages of the live cycle of robots. In particular we present how model-based testing and diagnosis can be used for this goal and how suitable models for these approaches can be obtained. The proposed approach was evaluated in a real industrial use-case showing superior performance compared to the hand-coded solutions used before.

Autonomous robots are artifacts that comprise a significant number of heterogeneous hardware and software components and interact with dynamic environments. Therefore, there is always a chance of faults at run-time that negatively affect the reliability of the system. In this paper we present a novel diagnosis and repair architecture for ROS-based robot systems. It is an extension to the existing ROS diagnostics stack and follows a model-based diagnosis and repair approach. In the paper we discuss the integrated diagnosis and repair architecture in detail. Moreover, we show its application to an example robot system and report first experimental results. The presented work provides three major contributions: a combination of diagnosis and repair, the integration of hardware and software, and the integration into ROS.

The tasks robots are employed to achieve are becoming increasingly complex, demanding for dependable operation, especially if robots and humans share common space. Unfortunately, for these robots non-determinism is a severe challenge. Malfunctioning hardware, inaccurate sensors, exogenous events and incomplete knowledge lead to inconsistencies in the robot's belief about the world.

The robot programming and plan language IndiGolog allows for on-line execution of actions and offline projections of programs in dynamic and partly unknown environments. Basic assumptions are that the outcomes of primitive and sensing actions are correctly modeled, and that the agent is informed about all exogenous events beyond its control. In real-world applications, however, such assumptions do not hold. In fact, an action's outcome is error-prone and sensing results are noisy. In this paper, we present a belief management system in IndiGolog that is able to detect inconsistencies between a robot's modeled belief and what happened in reality. The system furthermore derives explanations and maintains a consistent belief. Our main contributions are (1) a belief management system following a history-based diagnosis approach that allows an agent to actively cope with faulty actions and the occurrence of exogenous events; and (2) an implementation in IndiGolog and experimental results from a delivery domain.

The control system for a mobile robot typically assumes fixed kinematics according to the drive's geometry and functionality. Faults in the system, for example a blocked steering actuator, will then lead to an undesired behaviour, unless one takes care of specific single and/or multiple faults explicitly. We present a novel model-programmed procedure for on-line kinematics reasoning that allows a robot to deduce the (inverse)-kinematics of the drive and also its kinematic abilities for the specific modes of operation and some falt modes during operation. As a consequence, we can reconfigure a robot drive to compensate for some faults and also inform a higher level control system about changed mobility capabilities of a robot. Being fault tolerant is, however, only one advantage of our approach that derives the kinematics control strategy from a geometric and functional model of the drive. We can easily adapt the controller for various robot drives, handle drives that change their geometry and functionality during run-time and also provide the basis for a flexible control scheme for self-configuring multi-robot systems.

We have built an autonomous robot, out of standard components, and combined it with optimal game winning strategies. This results in an artificial companion which plays the board game Pylos in a fully interactive manner and up to the highest possible level.

Modern drives of mobile robots are complex machines. Because of this complexity, as well as of wear and aging of components, faults occurs in such systems quite frequently at runtime. In order to use such drives in truly autonomous robots it is desirable that the robot is able to automatically react to such faults. Therefore, the robot needs reasoning and reconfiguration capabilities in order to be able to detect, localize and repair such faults on-line. In this paper we propose a model-based diagnosis and reconfiguration framework which allows an autonomous robot to detect and compensate faults in its drive. Moreover, we present an implementation for a real robot platform. Finally, we report experimental results which shows that the proposed framework is able to correctly cope with injected faults in the drive hardware, like broken motors.