12 June 2020 in IEEE Sensors Journal
HAHojjat Gharaei Amirabadi
Biomechanical sensors are essential components for wearable robots because of their broad spectrum of industrial and medical applications. A wide range of these sensors uses soft materials with deformation-related electrical properties, namely variable resistors and capacitors. Previously, conductive materials were injected into silicone structures that require several molding steps to build the microchannels. This paper proposes a versatile soft sensing glove, using a simple process for preparing sensors of different sizes without molding. We used commercially available silicone tubes to host the conductive liquid. Ten sensors were attached to the back side of the glove to measure flexion-extension and four sensors were placed on the glove at the interdigital folds between the fingers to measure abduction-adduction. The sensory glove successfully replicates hand motion. We used machine learning algorithms to estimate the angles of the joints in the hand and also to identify 15 gestures. The system’s robustness was evaluated in two experiments. The gesture prediction is robust to shocks from contact with a punching ball and also submersion of the glove in water. The proposed sensory glove overcomes challenges of comfort, rigidity, and robustness. It can be used, therefore, to replicate human hand motion in industrial and medical applications.
27 July 2020 in IEEE Transactions on Medical Robotics and Bionics
Textile materials have recently been used in wearable robotics to provide more compact, lightweight, and low-cost solutions. Designing the textile-based actuator remains challenging in terms of providing a spectrum of forces that support the human body motion. We propose a new design for a textile-based actuator, which is made of an inflatable tube folded inside a housing fabric. The direction of the folds is perpendicular to the actuation plane allowing the actuator to exhibit high performance in dealing with heavy weights; it also demonstrates coherent behaviors with a broad set of its geometric parameters. We built a wearable glove that exhibits high performance in holding 9 kg weights. We show that with a larger scale, the bending actuator can hold up to 20 kg. Furthermore, a physical assistive experiment shows a reduction of 50% in electromyography signals reflecting flexor digitorum sublimis muscle activity during 9 kg weight bar grasping and releasing stages. Finally, we propose a variable curvature design, which provides a new perspective for the human-centered design of wearable robots.
21 December 2018 in IEEE Transactions on Systems, Man, and Cybernetics: Systems
We present an approach on fall detection and recovery perturbation during humanoid robot swinging. Reinforcement learning (Q-learning) is employed to explore relationship between actions and states that allow the robot to trigger a reaction to avoid falling. A self-organizing map (SOM) is employed using a circular topological neighborhood function to transform continuous exteroceptive information of the robot during stable swinging into a discrete representation of states. We take advantage of the SOM clustering and topology preservation in the perturbation detection. Swinging and recovery actions are generated from the same neural model using a multilayered multipattern central pattern generator. Experiments, which were carried out in the simulation and on the real humanoid robot (NAO), show that our approach allows humanoid robots to recover from pushing successfully by learning to switch from a rhythmic to an appropriate nonrhythmic behavior.
30 January 2019 in IEEE Transactions on Cognitive and Developmental Systems
How can robotics benefit from neuroscience to build a unified framework that computes actions for both, locomotion and manipulation tasks? Inspired by the hierarchical neural control of movement from cortex to spinal cord, we propose a model that generates a concrete action representation in robotics. The action program is composed of four basic modules: pattern selection, spatial coordination, temporal coordination, and sensory motor adaptation. The first and the fourth are considered for behaviour initiation. The model is implemented on a humanoid robot to generate rhythmic and non-rhythmic movements. The robot is able to perform tasks like perturbation recovery, and drawing based on different motor programs generated by the same model. Unifying motor control in robotics through a hierarchical structure increases the capacity to gain an accurate and deep understanding of transfer of motor skills between different tasks.
24 March 2019 in Sensors and Actuators A: Physical
23 July 2018 in Frontiers in Neurorobotics
How do robots learn to perform motor tasks in a specific condition and apply what they have learned in a new condition? This paper proposes a framework for motor coordination acquisition of a robot drawing straight lines within a part of the workspace. Then, it addresses transferring the acquired coordination into another area of the workspace while performing the same task. Motor patterns are generated by a Central Pattern Generator (CPG) model. The motor coordination for a given task is acquired by using a multi-objective optimization method that adjusts the CPGs' parameters involved in the coordination. To transfer the acquired motor coordination to the whole workspace we employed (1) a Self-Organizing Map that represents the end-effector coordination in the Cartesian space, and (2) an estimation method based on Inverse Distance Weighting that estimates the motor program parameters for each SOM neuron. After learning, the robot generalizes the acquired motor program along the SOM network. It is able therefore to draw lines from any point in the 2D workspace and with different orientations. Aside from the obvious distinctiveness of the proposed framework from those based on inverse kinematics typically leading to a point-to-point drawing, our approach also permits of transferring the motor program throughout the workspace.