Frank Kirchner

Raghav Soni, Daniel Harnack, Hannah Isermann et al.

Legged locomotion is arguably the most suited and versatile mode to deal with natural or unstructured terrains. Intensive research into dynamic walking and running controllers has recently yielded great advances, both in the optimal control and reinforcement learning (RL) literature. Hopping is a challenging dynamic task involving a flight phase and has the potential to increase the traversability of legged robots. Model based control for hopping typically relies on accurate detection of different jump phases, such as lift-off or touch down, and using different controllers for each phase. In this paper, we present a end-to-end RL based torque controller that learns to implicitly detect the relevant jump phases, removing the need to provide manual heuristics for state detection. We also extend a method for simulation to reality transfer of the learned controller to contact rich dynamic tasks, resulting in successful deployment on the robot after training without parameter tuning.

Amr Gomaa, Bilal Mahdy, Niko Kleer et al.

Recent advances in machine learning models allowed robots to identify objects on a perceptual nonsymbolic level (e.g., through sensor fusion and natural language understanding). However, these primarily black-box learning models still lack interpretation and transferability and require high data and computational demand. An alternative solution is to teach a robot on both perceptual nonsymbolic and conceptual symbolic levels through hybrid neurosymbolic learning approaches with expert feedback (i.e., human-in-the-loop learning). This work proposes a concept for this user-centered hybrid learning paradigm that focuses on robotic surgical situations. While most recent research focused on hybrid learning for non-robotic and some generic robotic domains, little work focuses on surgical robotics. We survey this related research while focusing on human-in-the-loop surgical robotic systems. This evaluation highlights the most prominent solutions for autonomous surgical robots and the challenges surgeons face when interacting with these systems. Finally, we envision possible ways to address these challenges using online apprenticeship learning based on implicit and explicit feedback from expert surgeons.

Franek Stark, Felix Wiebe, Shubham Vyas et al.

This paper presents a multi-phase whole-body model predictive control approach for bipedal walking, combining a detailed whole-body model in the near horizon with a simplified single-rigid-body model in the later prediction steps. This reduces computational complexity while retaining prediction capabilities. The resulting nonlinear optimal control problem is solved using sequential quadratic programming (SQP) in acados. Using a prior specified contact schedule and a target walking speed, the controller optimizes joint torques without depending on prior selected foot step locations. The controller is validated in MuJoCo simulation on the 18-DoF bipedal robot HyPer-2

Dennis Mronga, Andreas Bresser, Fabian Maas et al.

In this paper, we present the service robot MARLIN and its integration with the K4R platform, a cloud system for complex AI applications in retail. At its core, this platform contains so-called semantic digital twins, a semantically annotated representation of the retail store. MARLIN continuously exchanges data with the K4R platform, improving the robot's capabilities in perception, autonomous navigation, and task planning. We exploit these capabilities in a retail intralogistics scenario, specifically by assisting store employees in stocking shelves. We demonstrate that MARLIN is able to update the digital representation of the retail store by detecting and classifying obstacles, autonomously planning and executing replenishment missions, adapting to unforeseen changes in the environment, and interacting with store employees. Experiments are conducted in simulation, in a laboratory environment, and in a real store. We also describe and evaluate a novel algorithm for autonomous navigation of articulated tractor-trailer systems. The algorithm outperforms the manufacturer's proprietary navigation approach and improves MARLIN's navigation capabilities in confined spaces.

Octavio Arriaga, Jichen Guo, Rebecca Adam et al.

Humans excel at building generalizations of new concepts from just one single example. Contrary to this, current computer vision models typically require large amount of training samples to achieve a comparable accuracy. In this work we present a Bayesian model of perception that learns using only minimal data, a prototypical probabilistic program of an object. Specifically, we propose a generative inverse graphics model of primitive shapes, to infer posterior distributions over physically consistent parameters from one or several images. We show how this representation can be used for downstream tasks such as few-shot classification and pose estimation. Our model outperforms existing few-shot neural-only classification algorithms and demonstrates generalization across varying lighting conditions, backgrounds, and out-of-distribution shapes. By design, our model is uncertainty-aware and uses our new differentiable renderer for optimizing global scene parameters through gradient descent, sampling posterior distributions over object parameters with Markov Chain Monte Carlo (MCMC), and using a neural based likelihood function.

Yuhan Jin, Nayari Marie Lessa, Mariela De Lucas Alvarez et al.

Marine ecosystem degradation necessitates continuous, scientifically selective underwater monitoring. However, most autonomous underwater vehicles (AUVs) operate as passive data loggers, capturing exhaustive video for offline review and frequently missing transient events of high scientific value. Transitioning to active perception requires a causal, online signal that highlights significant phenomena while suppressing maneuver-induced visual changes. We propose DINO-Explorer, a novelty-aware perception framework driven by a continuous semantic surprise signal. Operating within the latent space of a frozen DINOv3 foundation model, it leverages a lightweight, action-conditioned recurrent predictor to anticipate short-horizon semantic evolution. An efference-copy-inspired module utilizes globally pooled optical flow to discount self-induced visual changes without suppressing genuine environmental novelty. We evaluate this signal on the downstream task of asynchronous event triage under variant telemetry constraints. Results demonstrate that DINO-Explorer provides a robust, bandwidth-efficient attention mechanism. At a fixed operating point, the system retains 78.8% of post-discovery human-reviewer consensus events with a 56.8% trigger confirmation rate, effectively surfacing mission-relevant phenomena. Crucially, ego-motion conditioning suppresses 45.5% of false positives relative to an uncompensated surprise signal baseline. In a replay-side Pareto ablation study, DINO-Explorer robustly dominates the validated peak F1 versus telemetry bandwidth frontier, reducing telemetry bandwidth by 48.2% at the selected operating point while maintaining a 62.2% peak F1 score, successfully concentrating data transmission around human-verified novelty events.

Melvin Laux, Yi-Ling Liu, Rina Alo et al.

Although autonomous underwater vehicles promise the capability of marine ecosystem monitoring, their deployment is fundamentally limited by the difficulty of controlling vehicles under highly uncertain and non-stationary underwater dynamics. To address these challenges, we employ a data-driven reinforcement learning approach to compensate for unknown dynamics and task variations.Traditional single-task reinforcement learning has a tendency to overfit the training environment, thus, limit the long-term usefulness of the learnt policy. Hence, we propose to use a contextual multi-task reinforcement learning paradigm instead, allowing us to learn controllers that can be reused for various tasks, e.g., detecting oysters in one reef and detecting corals in another. We evaluate whether contextual multi-task reinforcement learning can efficiently learn robust and generalisable control policies for autonomous underwater reef monitoring. We train a single context-dependent policy that is able to solve multiple related monitoring tasks in a simulated reef environment in HoloOcean. In our experiments, we empirically evaluate the contextual policies regarding sample-efficiency, zero-shot generalisation to unseen tasks, and robustness to varying water currents. By utilising multi-task reinforcement learning, we aim to improve the training effectiveness, as well as the reusability of learnt policies to take a step towards more sustainable procedures in autonomous reef monitoring.

Yi-Ling Liu, Melvin Laux, Mariela De Lucas Alvarez et al.

Autonomous underwater vehicles are required to perform multiple tasks adaptively and in an explainable manner under dynamic, uncertain conditions and limited sensing, challenges that classical controllers struggle to address. This demands robust, generalizable, and inherently interpretable control policies for reliable long-term monitoring. Reinforcement learning, particularly multi-task RL, overcomes these limitations by leveraging shared representations to enable efficient adaptation across tasks and environments. However, while such policies show promising results in simulation and controlled experiments, they yet remain opaque and offer limited insight into the agent's internal decision-making, creating gaps in transparency, trust, and safety that hinder real-world deployment. The internal policy structure and task-specific specialization remain poorly understood. To address these gaps, we analyze the internal structure of a pretrained multi-task reinforcement learning network in the HoloOcean simulator for underwater navigation by identifying and comparing task-specific subnetworks responsible for navigating toward different species. We find that in a contextual multi-task reinforcement learning setting with related tasks, the network uses only about 1.5% of its weights to differentiate between tasks. Of these, approximately 85% connect the context-variable nodes in the input layer to the next hidden layer, highlighting the importance of context variables in such settings. Our approach provides insights into shared and specialized network components, useful for efficient model editing, transfer learning, and continual learning for underwater monitoring through a contextual multi-task reinforcement learning method.

Daniel Harnack, Christoph Lüth, Lukas Gross et al.

Generating physical movement behaviours from their symbolic description is a long-standing challenge in artificial intelligence (AI) and robotics, requiring insights into numerical optimization methods as well as into formalizations from symbolic AI and reasoning. In this paper, a novel approach to finding a reward function from a symbolic description is proposed. The intended system behaviour is modelled as a hybrid automaton, which reduces the system state space to allow more efficient reinforcement learning. The approach is applied to bipedal walking, by modelling the walking robot as a hybrid automaton over state space orthants, and used with the compass walker to derive a reward that incentivizes following the hybrid automaton cycle. As a result, training times of reinforcement learning controllers are reduced while final walking speed is increased. The approach can serve as a blueprint how to generate reward functions from symbolic AI and reasoning.

Sara Khan, Mehmed Yüksel, Frank Kirchner

Wear and tear detection in fleet and shared vehicle systems is a critical challenge, particularly in rental and car-sharing services, where minor damage, such as dents, scratches, and underbody impacts, often goes unnoticed or is detected too late. Currently, manual inspection methods are the default approach, but are labour-intensive and prone to human error. In contrast, state-of-the-art image-based methods are less reliable when the vehicle is moving, and they cannot effectively capture underbody damage due to limited visual access and spatial coverage. This work introduces a novel multi-modal architecture based on anomaly detection to address these issues. Sensors such as Inertial Measurement Units (IMUs) and microphones are integrated into a compact device mounted on the vehicle's windshield. This approach supports real-time damage detection while avoiding the need for highly resource-intensive sensors. We developed multiple variants of multi-modal autoencoder-based architectures and evaluated them against unimodal and state-of-the-art methods. Our multi-modal ensemble model with pooling achieved the highest performance, with a Receiver Operating Characteristic-Area Under Curve (ROC-AUC) of 92%, demonstrating its effectiveness in real-world applications. This approach can also be extended to other applications, such as improving automotive safety. It can integrate with airbag systems for efficient deployment and help autonomous vehicles by complementing other sensors in collision detection.

Octavio Arriaga, Rebecca Adam, Melvin Laux et al.

Real-world robotic applications, from autonomous exploration to assistive technologies, require adaptive, interpretable, and data-efficient learning paradigms. While deep learning architectures and foundation models have driven significant advances in diverse robotic applications, they remain limited in their ability to operate efficiently and reliably in unknown and dynamic environments. In this position paper, we critically assess these limitations and introduce a conceptual framework for combining data-driven learning with deliberate, structured reasoning. Specifically, we propose leveraging differentiable physics for efficient world modeling, Bayesian inference for uncertainty-aware decision-making, and meta-learning for rapid adaptation to new tasks. By embedding physical symbolic reasoning within neural models, robots could generalize beyond their training data, reason about novel situations, and continuously expand their knowledge. We argue that such hybrid neuro-symbolic architectures are essential for the next generation of autonomous systems, and to this end, we provide a research roadmap to guide and accelerate their development.

Siddhant Shete, Dennis Mronga, Ankita Jadhav et al.

Anomaly detection deals with detecting deviations from established patterns within data. It has various applications like autonomous driving, predictive maintenance, and medical diagnosis. To improve anomaly detection accuracy, transfer learning can be applied to large, pre-trained models and adapt them to the specific application context. In this paper, we propose a novel framework for online-adaptive anomaly detection using transfer learning. The approach adapts to different environments by selecting visually similar training images and online fitting a normality model to EfficientNet features extracted from the training subset. Anomaly detection is then performed by computing the Mahalanobis distance between the normality model and the test image features. Different similarity measures (SIFT/FLANN, Cosine) and normality models (MVG, OCSVM) are employed and compared with each other. We evaluate the approach on different anomaly detection benchmarks and data collected in controlled laboratory settings. Experimental results showcase a detection accuracy exceeding 0.975, outperforming the state-of-the-art ET-NET approach.

Niklas Kueper, Kartik Chari, Judith Bütefür et al.

This paper presents a dataset containing recordings of the electroencephalogram (EEG) and the electromyogram (EMG) from eight subjects who were assisted in moving their right arm by an active orthosis device. The supported movements were elbow joint movements, i.e., flexion and extension of the right arm. While the orthosis was actively moving the subject's arm, some errors were deliberately introduced for a short duration of time. During this time, the orthosis moved in the opposite direction. In this paper, we explain the experimental setup and present some behavioral analyses across all subjects. Additionally, we present an average event-related potential analysis for one subject to offer insights into the data quality and the EEG activity caused by the error introduction. The dataset described herein is openly accessible. The aim of this study was to provide a dataset to the research community, particularly for the development of new methods in the asynchronous detection of erroneous events from the EEG. We are especially interested in the tactile and haptic-mediated recognition of errors, which has not yet been sufficiently investigated in the literature. We hope that the detailed description of the orthosis and the experiment will enable its reproduction and facilitate a systematic investigation of the influencing factors in the detection of erroneous behavior of assistive systems by a large community.

Hans Hohenfeld, Dirk Heimann, Felix Wiebe et al.

We utilize hybrid quantum deep reinforcement learning to learn navigation tasks for a simple, wheeled robot in simulated environments of increasing complexity. For this, we train parameterized quantum circuits (PQCs) with two different encoding strategies in a hybrid quantum-classical setup as well as a classical neural network baseline with the double deep Q network (DDQN) reinforcement learning algorithm. Quantum deep reinforcement learning (QDRL) has previously been studied in several relatively simple benchmark environments, mainly from the OpenAI gym suite. However, scaling behavior and applicability of QDRL to more demanding tasks closer to real-world problems e. g., from the robotics domain, have not been studied previously. Here, we show that quantum circuits in hybrid quantum-classic reinforcement learning setups are capable of learning optimal policies in multiple robotic navigation scenarios with notably fewer trainable parameters compared to a classical baseline. Across a large number of experimental configurations, we find that the employed quantum circuits outperform the classical neural network baselines when equating for the number of trainable parameters. Yet, the classical neural network consistently showed better results concerning training times and stability, with at least one order of magnitude of trainable parameters more than the best-performing quantum circuits. However, validating the robustness of the learning methods in a large and dynamic environment, we find that the classical baseline produces more stable and better performing policies overall.

Sadique Adnan Siddiqui, Lisa Gutzeit, Frank Kirchner

In this work, we investigate the influence of labeling methods on the classification of human movements on data recorded using a marker-based motion capture system. The dataset is labeled using two different approaches, one based on video data of the movements, the other based on the movement trajectories recorded using the motion capture system. The dataset is labeled using two different approaches, one based on video data of the movements, the other based on the movement trajectories recorded using the motion capture system. The data was recorded from one participant performing a stacking scenario comprising simple arm movements at three different speeds (slow, normal, fast). Machine learning algorithms that include k-Nearest Neighbor, Random Forest, Extreme Gradient Boosting classifier, Convolutional Neural networks (CNN), Long Short-Term Memory networks (LSTM), and a combination of CNN-LSTM networks are compared on their performance in recognition of these arm movements. The models were trained on actions performed on slow and normal speed movements segments and generalized on actions consisting of fast-paced human movement. It was observed that all the models trained on normal-paced data labeled using trajectories have almost 20% improvement in accuracy on test data in comparison to the models trained on data labeled using videos of the performed experiments.

Adrian Lubitz, Matias Valdenegro-Toro, Frank Kirchner

Robots are becoming everyday devices, increasing their interaction with humans. To make human-machine interaction more natural, cognitive features like Visual Voice Activity Detection (VVAD), which can detect whether a person is speaking or not, given visual input of a camera, need to be implemented. Neural networks are state of the art for tasks in Image Processing, Time Series Prediction, Natural Language Processing and other domains. Those Networks require large quantities of labeled data. Currently there are not many datasets for the task of VVAD. In this work we created a large scale dataset called the VVAD-LRS3 dataset, derived by automatic annotations from the LRS3 dataset. The VVAD-LRS3 dataset contains over 44K samples, over three times the next competitive dataset (WildVVAD). We evaluate different baselines on four kinds of features: facial and lip images, and facial and lip landmark features. With a Convolutional Neural Network Long Short Term Memory (CNN LSTM) on facial images an accuracy of 92% was reached on the test set. A study with humans showed that they reach an accuracy of 87.93% on the test set.

Tao Wang, Frank Kirchner

With an increasing demand for robots, robotic grasping will has a more important role in future applications. This paper takes grasp stability prediction as the key technology for grasping and tries to solve the problem with time series data inputs including the force and pressure data. Widely applied to more fields to predict unstable grasping with time series data, algorithms can significantly promote the application of artificial intelligence in traditional industries. This research investigates models that combine short-time Fourier transform (STFT) and long short-term memory (LSTM) and then tested generalizability with dexterous hand and suction cup gripper. The experiments suggest good results for grasp stability prediction with the force data and the generalized results in the pressure data. Among the 4 models, (Data + STFT) & LSTM delivers the best performance. We plan to perform more work on grasp stability prediction, generalize the findings to different types of sensors, and apply the grasp stability prediction in more grasping use cases in real life.

Julian Esser, Shivesh Kumar, Heiner Peters et al.

Last decades of humanoid research has shown that humanoids developed for high dynamic performance require a stiff structure and optimal distribution of mass--inertial properties. Humanoid robots built with a purely tree type architecture tend to be bulky and usually suffer from velocity and force/torque limitations. This paper presents a novel series-parallel hybrid humanoid called RH5 which is 2 m tall and weighs only 62.5 kg capable of performing heavy-duty dynamic tasks with 5 kg payloads in each hand. The analysis and control of this humanoid is performed with whole-body trajectory optimization technique based on differential dynamic programming (DDP). Additionally, we present an improved contact stability soft-constrained DDP algorithm which is able to generate physically consistent walking trajectories for the humanoid that can be tracked via a simple PD position control in a physics simulator. Finally, we showcase preliminary experimental results on the RH5 humanoid robot.

Matthias Rosynski, Frank Kirchner, Matias Valdenegro-Toro

Deep Reinforcement Learning (DRL) connects the classic Reinforcement Learning algorithms with Deep Neural Networks. A problem in DRL is that CNNs are black-boxes and it is hard to understand the decision-making process of agents. In order to be able to use RL agents in highly dangerous environments for humans and machines, the developer needs a debugging tool to assure that the agent does what is expected. Currently, rewards are primarily used to interpret how well an agent is learning. However, this can lead to deceptive conclusions if the agent receives more rewards by memorizing a policy and not learning to respond to the environment. In this work, it is shown that this problem can be recognized with the help of gradient visualization techniques. This work brings some of the best-known visualization methods from the field of image classification to the area of Deep Reinforcement Learning. Furthermore, two new visualization techniques have been developed, one of which provides particularly good results. It is being proven to what extent the algorithms can be used in the area of Reinforcement learning. Also, the question arises on how well the DRL algorithms can be visualized across different environments with varying visualization techniques.

Mohandass Muthuraja, Octavio Arriaga, Paul Plöger et al.

Object detectors have improved considerably in the last years by using advanced CNN architectures. However, many detector hyper-parameters are generally manually tuned, or they are used with values set by the detector authors. Automatic Hyper-parameter optimization has not been explored in improving CNN-based object detectors hyper-parameters. In this work, we propose the use of Black-box optimization methods to tune the prior/default box scales in Faster R-CNN and SSD, using Bayesian Optimization, SMAC, and CMA-ES. We show that by tuning the input image size and prior box anchor scale on Faster R-CNN mAP increases by 2% on PASCAL VOC 2007, and by 3% with SSD. On the COCO dataset with SSD there are mAP improvement in the medium and large objects, but mAP decreases by 1% in small objects. We also perform a regression analysis to find the significant hyper-parameters to tune.

Octavio Arriaga, Matias Valdenegro-Toro, Mohandass Muthuraja et al.

In this paper we introduce the Perception for Autonomous Systems (PAZ) software library. PAZ is a hierarchical perception library that allow users to manipulate multiple levels of abstraction in accordance to their requirements or skill level. More specifically, PAZ is divided into three hierarchical levels which we refer to as pipelines, processors, and backends. These abstractions allows users to compose functions in a hierarchical modular scheme that can be applied for preprocessing, data-augmentation, prediction and postprocessing of inputs and outputs of machine learning (ML) models. PAZ uses these abstractions to build reusable training and prediction pipelines for multiple robot perception tasks such as: 2D keypoint estimation, 2D object detection, 3D keypoint discovery, 6D pose estimation, emotion classification, face recognition, instance segmentation, and attention mechanisms.

Dennis Mronga, Frank Kirchner

Constraint-based control approaches offer a flexible way to specify robotic manipulation tasks and execute them on robots with many degrees of freedom. However, the specification of task constraints and their associated priorities usually requires a human-expert and often leads to tailor-made solutions for specific situations. This paper presents our recent efforts to automatically derive task constraints for a constraint-based robot controller from data and adapt them with respect to previously unseen situations (contexts). We use a programming-by-demonstration approach to generate training data in multiple variations (context changes) of a given task. From this data we learn a probabilistic model that maps context variables to task constraints and their respective soft task priorities. We evaluate our approach with 3 different dual-arm manipulation tasks on an industrial robot and show that it performs better in terms of reproduction accuracy than constraint-based controllers with manually specified constraints.

Thomas M. Roehr, Daniel Harnack, Hendrik Wöhrle et al.

In this paper we introduce Q-Rock, a development cycle for the automated self-exploration and qualification of robot behaviors. With Q-Rock, we suggest a novel, integrative approach to automate robot development processes. Q-Rock combines several machine learning and reasoning techniques to deal with the increasing complexity in the design of robotic systems. The Q-Rock development cycle consists of three complementary processes: (1) automated exploration of capabilities that a given robotic hardware provides, (2) classification and semantic annotation of these capabilities to generate more complex behaviors, and (3) mapping between application requirements and available behaviors. These processes are based on a graph-based representation of a robot's structure, including hardware and software components. A central, scalable knowledge base enables collaboration of robot designers including mechanical, electrical and systems engineers, software developers and machine learning experts. In this paper we formalize Q-Rock's integrative development cycle and highlight its benefits with a proof-of-concept implementation and a use case demonstration.

Swaroop Bhandary K, Nico Hochgeschwender, Paul Plöger et al.

Deep learning models are extensively used in various safety critical applications. Hence these models along with being accurate need to be highly reliable. One way of achieving this is by quantifying uncertainty. Bayesian methods for UQ have been extensively studied for Deep Learning models applied on images but have been less explored for 3D modalities such as point clouds often used for Robots and Autonomous Systems. In this work, we evaluate three uncertainty quantification methods namely Deep Ensembles, MC-Dropout and MC-DropConnect on the DarkNet21Seg 3D semantic segmentation model and comprehensively analyze the impact of various parameters such as number of models in ensembles or forward passes, and drop probability values, on task performance and uncertainty estimate quality. We find that Deep Ensembles outperforms other methods in both performance and uncertainty metrics. Deep ensembles outperform other methods by a margin of 2.4% in terms of mIOU, 1.3% in terms of accuracy, while providing reliable uncertainty for decision making.

Felix Wiebe, Shivesh Kumar, Daniel Harnack et al.

Motion planning is a difficult problem in robot control. The complexity of the problem is directly related to the dimension of the robot's configuration space. While in many theoretical calculations and practical applications the configuration space is modeled as a continuous space, we present a discrete robot model based on the fundamental hardware specifications of a robot. Using lattice path methods, we provide estimates for the complexity of motion planning by counting the number of possible trajectories in a discrete robot configuration space.

Tim von Oehsen, Alexander Fabisch, Shivesh Kumar et al.

Several publications are concerned with learning inverse kinematics, however, their evaluation is often limited and none of the proposed methods is of practical relevance for rigid-body kinematics with a known forward model. We argue that for rigid-body kinematics one of the first proposed machine learning (ML) solutions to inverse kinematics -- distal teaching (DT) -- is actually good enough when combined with differentiable programming libraries and we provide an extensive evaluation and comparison to analytical and numerical solutions. In particular, we analyze solve rate, accuracy, sample efficiency and scalability. Further, we study how DT handles joint limits, singularities, unreachable poses, trajectories and provide a comparison of execution times. The three approaches are evaluated on three different rigid body mechanisms with varying complexity. With enough training data and relaxed precision requirements, DT has a better solve rate and is faster than state-of-the-art numerical solvers for a 15-DoF mechanism. DT is not affected by singularities while numerical solutions are vulnerable to them. In all other cases numerical solutions are usually better. Analytical solutions outperform the other approaches by far if they are available.

Alexander Fabisch, Christoph Petzoldt, Marc Otto et al.

Recent success of machine learning in many domains has been overwhelming, which often leads to false expectations regarding the capabilities of behavior learning in robotics. In this survey, we analyze the current state of machine learning for robotic behaviors. We will give a broad overview of behaviors that have been learned and used on real robots. Our focus is on kinematically or sensorially complex robots. That includes humanoid robots or parts of humanoid robots, for example, legged robots or robotic arms. We will classify presented behaviors according to various categories and we will draw conclusions about what can be learned and what should be learned. Furthermore, we will give an outlook on problems that are challenging today but might be solved by machine learning in the future and argue that classical robotics and other approaches from artificial intelligence should be integrated more with machine learning to form complete, autonomous systems.

Bilal Wehbe, Marc Hildebrandt, Frank Kirchner

Learning the dynamics of robots from data can help achieve more accurate tracking controllers, or aid their navigation algorithms. However, when the actual dynamics of the robots change due to external conditions, on-line adaptation of their models is required to maintain high fidelity performance. In this work, a framework for on-line learning of robot dynamics is developed to adapt to such changes. The proposed framework employs an incremental support vector regression method to learn the model sequentially from data streams. In combination with the incremental learning, strategies for including and forgetting data are developed to obtain better generalization over the whole state space. The framework is tested in simulation and real experimental scenarios demonstrating its adaptation capabilities to changes in the robot's dynamics.

Bilal Wehbe, Octavio Arriaga, Mario Michael Krell et al.

Multi-context model learning is crucial for marine robotics where several factors can cause disturbances to the system's dynamics. This work addresses the problem of identifying multiple contexts of an AUV model. We build a simulation model of the robot from experimental data, and use it to fill in the missing data and generate different model contexts. We implement an architecture based on long-short-term-memory (LSTM) networks to learn the different contexts directly from the data. We show that the LSTM network can achieve high classification accuracy compared to baseline methods, showing robustness against noise and scaling efficiently on large datasets.