Search

This paper describes a deep reinforcement learning (DRL) approach that won Phase 1 of the Real Robot Challenge (RRC) 2021, and then extends this method to a more difficult manipulation task. The RRC consisted of using a TriFinger robot to manipulate a cube along a specified positional trajectory, but with no requirement for the cube to have any specific orientation. We used a relatively simple reward function, a combination of a goal‐based sparse reward and a distance reward, in conjunction with Hindsight Experience Replay (HER) to guide the learning of the DRL agent (Deep Deterministic Policy Gradient [DDPG]). Our approach allowed our agents to acquire dexterous robotic manipulation strategies in simulation. These strategies were then deployed on the real robot and outperformed all other competition submissions, including those using more traditional robotic control techniques, in the final evaluation stage of the RRC. Here we extend this method, by modifying the task of Phase 1 of the RRC to require the robot to maintain the cube in a particular orientation, while the cube is moved along the required positional trajectory. The requirement to also orient the cube makes the agent less able to learn the task through blind exploration due to increased problem complexity. To circumvent this issue, we make novel use of a Knowledge Transfer (KT) technique that allows the strategies learned by the agent in the original task (which was agnostic to cube orientation) to be transferred to this task (where orientation matters). KT allowed the agent to learn and perform the extended task in the simulator, which improved the average positional deviation from 0.134 to 0.02 m, and average orientation deviation from 142° to 76° during evaluation. This KT concept shows good generalization properties and could be applied to any actor‐critic learning algorithm. [ABSTRACT FROM AUTHOR]

Blockchain designed for Mobile Ad hoc Networks (MANETs) and mesh networks is an emerging research topic that has to cope with the network partition problem. However, existing consensus algorithms used in blockchain have been designed to work in a fully connected network with reliable communication. As this assumption does not hold anymore in mobile wireless networks, we describe in this paper the problem of network partitions and their impact on blockchain. Then, we propose a new consensus algorithm called Consensus for Mesh (C4M) which is inspired by RAFT as a solution to this problem. The C4M consensus algorithm is integrated with Blockgraph, a blockchain solution for MANET and mesh networks. We implemented our solution in NS-3 to analyze its performances through simulations. The simulation results gave the first characterization of our algorithm, its performance, and its limits, especially in case of topology changes. [ABSTRACT FROM AUTHOR]

The static coefficient of friction (µstatic) plays an important role in dexterous object manipulation. Minimal normal force (i.e., grip force) needed to avoid dropping an object is determined by the tangential force at the fingertip-object contact and the frictional properties of the skin-object contact. Although frequently assumed to be constant for all levels of normal force (NF, the force normal to the contact), µ static actually varies nonlinearly with NF and increases at low NF levels. No method is currently available to measure the relationship between µstatic and NF easily. Therefore, we propose a new method allowing the simple and reliable measurement of the fingertip µstatic at different NF levels, as well as an algorithm for determining µstatic from measured forces and torques. Our method is based on active, back-and-forth movements of a subject's finger on the surface of a fixed six-axis force and torque sensor. µstatic is computed as the ratio of the tangential to the normal force at slip onset. A negative power law captures the relationship between µstatic and NF. Our method allows the continuous estimation of µstatic as a function of NF during dexterous manipulation, based on the relationship between µstatic and NF measured before manipulation. [ABSTRACT FROM PUBLISHER]

The article focuses on the role of home health aides in the management care of dying patients. The home health aides identified themselves as guests in the homes of the patients. Initially, most of them were not employed to tend dying patients, but they would be in service to them for an extended period and eventually would care for them when death was imminent. The aide and the patient provided one another with solace, comfort, and support. The aides expressed no overpowering need to dissociate themselves from an unpleasant situation but rather, a sense that they were honored to provide a helping hand for the one facing death. The home health aide sees death and the dying process as an unavoidable, natural condition that will eventually prevail over life. This attitude is conveyed in the work situation, as the aide works as an employee of the dying patient and the family. The home health aide looks on death not as a condition to be avoided or denied but, rather, one that must be met and interacted with, in a caring manner for as long as necessary until it has its ways.