Search Results (4)

Humanoid manipulators with kinematic redundancy offer enhanced dexterity and adaptability to complex environments. Solving their inverse kinematics (IK) is fundamental to trajectory tracking, motion planning, and real-time control. Conventional Jacobian-based iterative methods are widely used, but they are often sensitive to the initial guess, computationally expensive, and less effective in handling strict constraints. Arm-angle-based analytical parameterization reduces redundancy resolution to a single parameter. However, joint limits may lead to multiple disconnected feasible arm-angle intervals. Many existing methods still depend on a numerical search or intelligent optimization to select the arm angle. This lowers computational efficiency and provides less explicit control over branch and configuration selection. To address these issues, this paper extends the arm-angle analytical IK framework. It introduces global configuration parameters to explicitly control the shoulder-elbow-wrist configuration. It also completes the analytical derivation of the rotational relationships of the first three joints in the reference plane. In addition, a feasibility determination and modeling scheme for the arm-angle domain is established, which covers disconnected feasible intervals. The IK problem is then reformulated as a one-dimensional optimization over the feasible domain. An efficient interval-based search is employed to determine the optimal arm angle. Experimental results demonstrate high accuracy and interference-free trajectory tracking. Comparative tests on randomly sampled target poses are also performed. The results show more concentrated error distributions, shorter average computation time, and higher success rates. These results confirm the advantages of the proposed method in accuracy, robustness, and real-time performance. Full article

The lack of intuitive controllability remains a primary challenge in enabling transhumeral amputees to control a prosthesis for arm reaching with residual limb kinematics. Recent advancements in prosthetic arm control have focused on leveraging the predictive capabilities of artificial neural networks (ANNs) to automate elbow joint motion and wrist pronation–supination during target reaching tasks. However, large quantities of human motion data collected from different subjects for various activities of daily living (ADL) tasks are required to train these ANNs. For example, the reaching motion can be altered when the height of the desk is changed; however, it is cumbersome to conduct human experiments for all conditions. This paper proposes a framework for cloning motion datasets using deep reinforcement learning (DRL) to cater to training data requirements. DRL algorithms have been demonstrated to create human-like synergistic motion in humanoid agents to handle redundancy and optimize movements. In our study, we collected real motion data from six individuals performing multi-directional arm reaching tasks in the horizontal plane. We generated synthetic motion data that mimicked similar arm reaching tasks by utilizing a physics simulation and DRL-based arm manipulation. We then trained a CNN-LSTM network with different configurations of training motion data, including DRL, real, and hybrid datasets, to test the efficacy of the cloned motion data. The results of our evaluation showcase the effectiveness of the cloned motion data in training the ANN to predict natural elbow motion accurately across multiple subjects. Furthermore, motion data augmentation through combining real and cloned motion datasets has demonstrated the enhanced robustness of the ANN by supplementing and diversifying the limited training data. These findings have significant implications for creating synthetic dataset resources for various arm movements and fostering strategies for automatized prosthetic elbow motion. Full article

The article presents an algorithm for the multi-domain visual recognition of an indoor place. It is based on a convolutional neural network and style randomization. The authors proposed a scene classification mechanism and improved the performance of the models based on synthetic and real data from various domains. In the proposed dataset, a domain change was defined as a camera model change. A dataset of images collected from several rooms was used to show different scenarios, human actions, equipment changes, and lighting conditions. The proposed method was tested in a scene classification problem where multi-domain data were used. The basis was a transfer learning approach with an extension style applied to various combinations of source and target data. The focus was on improving the unknown domain score and multi-domain support. The results of the experiments were analyzed in the context of data collected on a humanoid robot. The article shows that the average score was the highest for the use of multi-domain data and data style enhancement. The method of obtaining average results for the proposed method reached the level of 92.08%. The result obtained by another research team was corrected. Full article

As an essential part for the development of autonomous agricultural robotics, the functional safety of autonomous agricultural machines is largely based on the functionality and robustness of non-contact sensor systems for human protection. This article presents a new step in the development of autonomous agricultural machine with a concept and the realization of a novel test method using a dynamic test stand on an agricultural farm in outdoor areas. With this test method, commercially available sensor systems are tested in a long-term test around the clock for 365 days a year and 24 h a day on a dynamic test stand in continuous outdoor use. A test over a longer period of time is needed to test as much as possible all occurring environmental conditions. This test is determined by the naturally occurring environmental conditions. This fact corresponds to the reality of unpredictable/determinable environmental conditions in the field and makes the test method and test stand so unique. The focus of the developed test methods is on creating own real environment detection areas (REDAs) for each sensor system, which can be used to compare and evaluate the autonomous human detection of the sensor systems for the functional safety of autonomous agricultural robots with a humanoid test target. Sensor manufacturers from industry and the automotive sector provide their sensor systems to have their sensors tested in cooperation with the TÜV. Full article