Deutsches Forschungszentrum für Künstliche Intelligenz GmbH (DFKI)
Lade...
Country
DE
City
Bremen
8 Ergebnisse
Filter
Einstellungen
Gerade angezeigt 1 - 8 von 8
- Some of the metrics are blocked by yourconsent settings
Item-typ:Veröffentlichung, Extending actuator-level control for robotic systems using distributed dynamics computation and incremental learning(2025-11-24); ; Actuators are fundamental components in robotic systems such as manipulator arms, since they enable a system to actively and physically interact with its environment. In classical industrial production environments, the focus of the optimization of this interaction has been precision and speed. The individual operating steps are typically tailored to a certain stage of the production and are hard programmed. Further interactions – including unintentional – are not considered within this scope and are therefore prevented by strict separation of workspaces for safety reasons. However, such limitations cannot be enforced if robotic systems are also to be used outside such defined workspaces. This is inevitably the case when they directly support or assist human individuals and thus share their workspace with humans, as is increasingly the case in industrial assembly processes, the service sector, and for assistance in the field of rehabilitation and household. In such cases, rigid position control of the joints is not sufficient, as it regards any deviations, irrespective of the current situation, as an error and increases the actuation forces in response. It does not take into account what driving forces are actually reasonable and necessary, nor whether a new situation has arisen that is causing the deviations. The aim of this dissertation is to extend the capabilities of the actuator-level control so that it incorporates the relationships between the generated forces and torques with the motion and external forces and learns to recognize new situations from its own sensory data. For this purpose, this work combines three approaches that build on each other and examines them experimentally. Firstly, methods and software tools are being developed that facilitate the creation of dynamic models based on physical insights in such a way that the missing parameter information can be obtained more easily from experimental data. The experimentally determined models in combination with an advanced motor control form the basis for safe compliant motion control of a manipulator arm without additional external sensors. Secondly, based on this, a method is developed that incorporates the dynamics locally at the actuator level by means of a distributed computation. This makes it possible to design more modular robot systems and to relieve a central computer, as well as to reduce the dependency on the data communication and associated latency that would normally be necessary for a central calculation. Thirdly, based on the concept of Adaptive Resonance Theory, a kind of episodic memory for recognizing different situations is developed. To achieve this, sensor data from robotic actuators are pre-processed in frequency domain and continuously – incrementally – learnt by an artificial neural network. It is shown experimentally that a robot system can distinguish collisions during an arm movement or jamming of parts during an assembly task from previous undisturbed executions.Dissertation34 51 - Some of the metrics are blocked by yourconsent settings
Item-typ:Veröffentlichung, Contact Implicit Control for the Vertical Hopper(2025-09-02); Hering-Bertram, MartinLegged robots offer significant potential for tasks in challenging environ- ments, but their control remains complex due to highly nonlinear dy- namics and hybrid contact behaviors. Model Predictive Control (MPC) is a promising approach, yet traditional methods often rely on predefined contact modes, sacrificing optimality and limiting adaptability. This the- sis explores a contact-implicit MPC framework for a vertical hopper to enable autonomous contact discovery and achieve desired jump heights without pre-planned trajectories. The developed controller is iLQR based and uses a relaxed contact model. It was successfully validated on an ex- ternal MuJoCo simulation, demonstrating its ability to generate smooth jumping motions. However, it could be observed that the relaxation used in the contact model introduced inaccuracies, particularly in impulse cal- culation, which led to performance degradation with larger timesteps and violated strict Signorini conditions. Recommendations for future work include implementing more sophisticated optimization methods like the bisection method, employing higher-order numerical integrators such as Runge-Kutta, and optimizing computational efficiency through matrix caching. This research underscores the functionality of contact-implicit control while identifying critical areas for improvement to achieve robust, real-time performance for dynamic legged systems.text::thesis::bachelor thesis85 68 - Some of the metrics are blocked by yourconsent settings
Item-typ:Veröffentlichung, PhysWM: Physical World Models for Robot LearningWithin the last decade machine learning methods have shown remarkable results in pattern recognition tasks and behavior learning. However, when applied to real-world robotics tasks, these approaches have limitations, such as sample inefficiency and limited generalization to out-of-distribution samples. Despite the availability of precise physics in simulation engines, model-based reinforcement learning (RL) resorts to learning an approximation of these dynamics. On the other hand, optimal control approaches often assume a static, complete model of the world, addressing the simulation-reality gap by adding low level controllers. In order to handle these issues, we propose a hybrid simulator consisting of differentiable physics and rendering modules, which employ symbolic representations and reduce the model complexity of neural policies, while retaining gradient computation for model and behavior optimization. Moreover, this reduced parametric representation enables the use of Bayesian inference to estimate the uncertainty over physical parameters. This uncertainty quantification allows us to generate a curriculum of exploration behaviors for continuously improving the world model.Konferenzbeitrag73 24 - Some of the metrics are blocked by yourconsent settings
Item-typ:Veröffentlichung, Advancements in parallel actuation: modeling, design, and applications(2025-08-27); ; ; Parallel actuation is both a solution and a challenge. Compared to serially actuated systems, it offers significant advantages in speed, accuracy, and efficiency. These improvements arise from transmitting motion through multiple mechanical branches simultaneously. However, this same characteristic also introduces new challenges at two distinct levels. First, the usable work-space is often significantly reduced, leading to designs that are highly task-specific. Second, the complexity of modeling increases, which has historically hindered the widespread adoption of optimal control algorithms compared to serial architectures. Furthermore, modern demands for increased robot compliance — to enhance adaptability, energy efficiency, and safety — stand in contrast to the inherently rigid nature of traditional parallel robots. This work addresses these challenges through various approaches. In the first part, belt-driven designs that lie at the intersection of serial and parallel architectures are explored. It is demonstrated that such systems enable high-speed movements within large work-spaces, thereby applying trajectory optimization methods. Additionally, the possibility to extend the usable work-space of parallel kinematic machines is exemplified. The second part focuses on compliant parallel actuation, showing that stiffness control can be achieved with a minimal number of actuators. It also introduces a novel concept combining soft-shell actuation with parallel linkages. The third part investigates various dynamic modeling techniques for parallel kinematic machines and provides a comparative analysis of their suitability for trajectory optimization.Dissertation47 47 - Some of the metrics are blocked by yourconsent settings
Item-typ:Veröffentlichung, Modular task modeling and creation of high-level building blocks for implementing complex robot applications(2025-06-24); ; Component-based frameworks for the development of robot control software provide tools to support software development and define a common component interface for the software created in this framework. This has a significant impact on robot programming: while the tools increase developer productivity, the common component interface increases software reusability. As a result, roboticists can now benefit from an extensive collection of software components. In a component-based software framework, the robot's behavior is defined by the selection of individual components, their configuration and the connections between the components. The reuse of software components significantly reduces development work, particularly in the development of general algorithms. For robot systems that have to be able to handle a wide range of applications with numerous behaviors and algorithmic approaches, current software frameworks lack the flexibility to ensure the necessary variability in robot control, which hinders the realization of versatile robot systems. Model-driven software development enables the realization of development processes in which recurring development tasks are automated through model transformation pipelines. The use of domain-specific languages makes it possible to design complex software systems in a more efficient and accessible way. This work investigates methods for the model-driven development of robot controllers. The focus is on the specification and execution of modular robot applications that combine a variety of different control paradigms and algorithms and can be adapted to different robot systems or application scenarios. A software ecosystem is presented, consisting of (a) a software system for modeling and executing robot controllers that are freely reconfigurable and interchangeable at the robot's runtime, (b) a high-level task modeling and execution environment that allows the integration of multiple behavior description paradigms into a common framework, and (c) a meta-modeling environment for creating high-level integration building blocks that can be adapted to different robot systems or tasks. After an introduction to the general research area and the technological foundations, the individual proposed technologies are presented. The underlying model representations and domain-specific languages are explained and the implementation of the software systems based on them is presented. Each software system is evaluated and discussed individually, whereby practical validation tests with real robot systems, dedicated performance measurements and theoretical considerations regarding the limits and performance of the software are made The evaluation confirms the applicability of the software for the intended application area of reconfigurable robot controllers and for the integration of complex control systems. In addition, open issues are identified and extensions of the proposed software are discussed for possible future work.Dissertation43 40 - Some of the metrics are blocked by yourconsent settings
Item-typ:Veröffentlichung, The Downgrading Axioms Challenge for Qualitative Composition of Food IngredientsQualitatively graded relations provide increased granularity for fine-grain modelling, and achieve qualitative abstraction from quantitative data. We focus on composite dosage for food ingredients in the BAALL Ontology (of considerable size): weight ratios, Alcohol By Volume, etc. To deduce the overall qualitative composition by reasoning, axioms for downgrading are introduced. These impose a heavy load on reasoning such that conventional reasoners fail.Konferenzbeitrag19 20 - Some of the metrics are blocked by yourconsent settings
Item-typ:Veröffentlichung, Design and field testing of a rover with an actively articulated suspension system in a Mars analog terrainThis study presents the electromechanical design, the control approach, and the results of a field test campaign with the hybrid wheeled-leg rover SherpaTT. The rover ranges in the 150 kg class and features an actively articulated suspension system comprising four legs with actively driven and steered wheels at each leg’s end. Five active degrees of freedom are present in each of the legs, resulting in 20 active degrees of freedom for the complete locomotion system. The control approach is based on force measurements at each wheel mounting point and roll–pitch measurements of the rover’s main body, allowing active adaption to sloping terrain, active shifting of the center of gravity within the rover’s support polygon, active roll–pitch influencing, and body-ground clearance control. Exteroceptive sensors such as camera or laser range finder are not required for ground adaption. A purely reactive approach is used, rendering a planning algorithm for stability control or force distribution unnecessary and thus simplifying the control efforts. The control approach was tested within a 4-week field deployment in the desert of Utah. The results presented in this paper substantiate the feasibility of the chosen approach: The main power requirement for locomotion is from the drive system, active adaption only plays a minor role in power consumption. Active force distribution between the wheels is successful in different footprints and terrain types and is not influenced by controlling the body’s roll–pitch angle in parallel to the force control. Slope-climbing capabilities of the system were successfully tested in slopes of up to 28° inclination, covered with loose soil and duricrust. The main contribution of this study is the experimental validation of the actively articulated suspension of SherpaTT in conjunction with a reactive control approach. Consequently, hardware and software design as well as experimentation are part of this study.Wissenschaftlicher Artikel169 287 - Some of the metrics are blocked by yourconsent settings
Item-typ:Veröffentlichung, The BAALL Ontology – Configuration of Service Robots, Food, and DietThe BAALL Ontology, originally motivated by Ambient Assisted Living, now comprises more than 40k OWL axioms to integrate diverse applications, covering a foundational, a variety of general, and several application domain ontologies for configuration of service robots, diets, structured food products and dishes, and cooking assistance. To maintain structural consistency, safe ontology extension is supported by Generic Ontology Design Patterns.Konferenzbeitrag18 30