Robots with high Degrees of Freedom (DoF) such as humanoids and mobile manipulators are expected to perform multiple tasks simultaneously. Hierarchical Quadratic Programming (HQP) can effectively compute a solution for strictly prioritized tasks. However, the continuity of control input is not guaranteed when the priorities of the tasks are modified during operation. This paper proposes a continuous task transition method for HQP based controller to insert, remove, and swap arbitrary tasks without discontinuity. Smooth task transition is assured because our approach uses activation parameters of the new and existing tasks without modifying control structure.The proposed approach is applied to various task transition scenarios including joint limit, singularity, and obstacle avoidance to guarantee the stable execution of the robot. The proposed control scheme was implemented on a 7-DoF robotic arm, and its performance was demonstrated by the continuity of control input during various task transition scenarios.

Recently, the robotic assembly has been expanded into an unstructured environment. This environment includes uncertainties that may cause unexpected situations such as a failure of the assembly. Such problems can be prevented or monitored by a robust contact state (CS) estimation method. In that sense, the paper suggests a CS estimation method that contains a torque indicator, a position/velocity indicator, and a CS discriminator. Using joint torque of manipulators and position/velocity of the end-effector, a Gaussian Mixture Model (GMM) builds each indicator by reflecting on two properties of measured data, i.e., non-stationary behavior and correlation among the data. The indicators play a role to indicate the corresponding sensor state. The discriminator is defined by rules which combine the results of the indicators, allowing a robust CS estimation to be achieved. In this respect, the proposed method has a distinct advantage over existing distance-based clustering methods which ignore probabilistic properties or correlation among measured data. The performance of the estimation is demonstrated through experiments with torque-controlled manipulators and commercial prefabricated furniture.

Multiarm systems can perform complex and difficult tasks, such as manipulating a heavy or large object, that cannot be accomplished by a single manipulator owing to workspace and payload limitations. However, the motion planning problem for performing such a task is challenging because of the need to consider the closed-chain constraint. This article proposes an efficient motion planner that considers the closed-chain constraint based on a probabilistic roadmap. The proposed planner utilizes the following strategies. First, the planner obtains feasible nodes by randomly sampling the object pose, followed by computing the inverse kinematics (IK) solution of the multiarm. This can directly find a collision-free node satisfying the closed-chain constraint. Second, the planner repeatedly updates the new IK solution of the multiarm for the start and goal object pose. The IK solution is computed as close as possible to the joint configuration of the neighbor node. Consequently, the planner is more efficient than the existing methods that generate a node by sampling the joint configuration with projection method and have one pair of the start and goal node. Therefore, the planner can efficiently compute the path for object manipulation using a multiarm under a closed-chain constraint. The effectiveness of the proposed planner is validated by comparison with the existing planners in several scenarios. A video clip of the experiments in various scenarios can be found at https://youtu.be/PR9aFf3juu4 .

Multi-axis actuation robotics systems comprising many joints with floating-base require more stability and safety than fixed robots and thus require more considerations. In this paper, the implementation of a real-time EtherCAT control system for the TOCABI humanoid robot with 33 degrees of freedom (DOF) is described. The focus is on the development of a high-performing EtherCAT master, which enables control of the robot’s high DOF at fast communication cycles. It also explores the use of a dual-channel EtherCAT master as a redundancy mechanism to handle communication disruptions and shows that this configuration reduces the burden on the communication network and increases the communication cycle, leading to good real-time performance. To show the advantages of the system, it examines the performance of the EtherCAT communication and evaluates the impact of RTnet on real-time performance, demonstrating that a high-performing EtherCAT master with hard real-time capabilities can be established using open-source software. The results of this work demonstrate the potential of using dual channels in EtherCAT master configurations and utilizing open-source software to implement low-cost EtherCAT master systems. The paper’s contribution to the field is its indication of developing stable robot systems with high DOF, which require hard real-time capability, even with open-source software.

This article presents a novel approach to address high-dimensional constrained motion planning problems. The proposed method exploits learned latent spaces to efficiently find a feasible constrained path. Although recent data-driven methods have reduced planning time, two major challenges arise as the space dimensionality increases, such as in multi-arm manipulation. First, the configuration space search employed by existing data-driven approaches becomes computationally expensive as the complexity of constraints increases. Second, preparing datasets for high-dimensional problems is a time-consuming task. To address these challenges, this article introduces a novel approach: the latent motion method. Instead of exploring the configuration space, the latent motion method explores the latent space with a latent jump method that mitigates topological problems. In addition, a tangent space dataset augmentation technique is employed to approximate the manifold using a sparse dataset. Experimental evaluations on benchmark problems indicate that the proposed approach outperforms existing methods in high-dimensional constrained motion planning problems.