The soft robotics field offers a large variety of actuators and robots, complex structures and designs able to behave in a required way, offering speed and resistance and at the same time reducing the contact damage chance. The control of such structures using mathematical methods requires complex methods which evolved over time in the form of predictive control and even neural networks and genetic algorithms. The current paper presents ongoing research that can be found at the intersection of meta-materials and sensor technology. It approaches a simpler method of controlling a flexible actuator based on a shape memory alloy, the regulator being a two-state switch controller steered by a pressure force sensor. We propose a complete analogical electrical structure and analyze the performance of the outcome, focusing on time response. The present work also contains a brief description of a mechanical design proposal able to optimize the impact of the shape memory alloy used, in order to obtain an amplified actuation.

Learning by demonstration, human-to-robot skills transfer or visual perception and control are tasks that can be tackled with techniques that combine machine vision and deep learning. In this paper, we take a step further from the classic deep learning based on the special Euclidean group (SE3) parameterization of displacement and motion. Considering an isomorphism between SE3 and the orthogonal dual tensor group, we propose an architecture that is set to improve the accuracy of computing the motion parameters. Thus, we introduce orthogonal dual tensors - convolutional neural network (${\widehat {{\mathbf{S}}\mathbb{O}}_3}$-CNN), a deep neural network set to learn from depth data how to extract semantic information regarding the layout of crobot’s segments while recovering the rigid displacement for each segment. To validate the design of ${\widehat {{\mathbf{S}}\mathbb{O}}_3}$-CNN, we employ a state-of-the-art database. The experimental results show that rigid displacements are estimated with higher accuracy by ${\widehat {{\mathbf{S}}\mathbb{O}}_3}$-CNN in comparison with a state-of-the-art SE3-CNN.

In recent years, quadcopters have become popular research platforms in the fields of robotics and control engineering. As a response to the academic community's needs, several open-source hardware and software platforms have emerged with the aim to impose as de-facto development tools. One of these platforms is represented by Crazyflie 2.0, a nano class quadcopter that was used in research projects that span from control architecture designs to high-level swarm synchronization.This paper presents a study of the control architecture employed in the stock firmware. Based on the mathematical model, the dynamic behaviour is thoroughly evaluated to understand the deviation between ideal and real evolution. A comprehensive phase of parameters tuning is conducted and multiple scenarios for real-time testing are employed. These scenarios comply in tracking predefined patterns, the quadcopter's evolution being tracked via a RGB-D sensor.The proposed evaluation procedure can be easily replicated and represents a very important step in obtaining accurate real-time results.

During the last few years, the field of micro aerial vehicles (drones) has encountered a significant focus among the robotics research community. Autonomous maneuvering for indoor environments is highly challenging on one hand due to space topology that can include different storage facilities, and on the other hand due to need for optimal planning that saves drone battery and task accomplishment time. In this paper we study a type of vehicle routing problem applied to an indoor warehouse. In particular, some goods should be transported from storage areas to delivery areas by limited energy drones with finite capacities for transporting goods. The presentation includes a planning solution based on mathematical programming and presents supporting simulations and real-time experiments.

The control of human hands and fingers represents one of the most complex functions of the motor cortex. In order to implement anthropomorphic hands that mimic accurately the motion ability of the human hands, the basic biological mechanisms of the natural muscle control should be modelled. This paper presents the design of a significantly improved control system based on analogue neural networks that can be used to replicate the biological control mechanisms of the natural muscles. In order to demonstrate the proposed concept, experiments were performed using a single-joint robotic arm that can be flexed as the human elbow by an artificial muscle connected as the biceps. In order to bring more biological plausibility to the robotic arm, the artificial muscle is implemented using a shape memory alloy wire which actuates by contraction as the natural muscles. Moreover, the contraction force of the actuator wire is directly determined by the spiking frequency of the electronic neurons as the motor neurons determines the contraction strength of the natural muscles. The experimental results show that using excitatory neurons and several inhibitory neurons unevenly distributed on the inputs of the artificial motor neurons that drive the shape memory alloy actuator, the spiking neural network is able to control with high precision the rotation of the arm mobile lever to random target positions even if the arm is slightly loaded. These results validate the control method of the robotic arm junction showing that the analogue spiking neural networks represents a very good alternative to control the contraction of SMA actuators in a biological plausible manner.

In this paper, a model predictive control strategy is presented to visual servoing a robot manipulator with eye-in-hand configuration. Starting with the discrete form of the time derivative of features related with camera velocity through image Jacobian matrix and taking into account the discrete robot model, a new approach for computing the predictions is introduced. By means of a cost function based on errors in image plane, convergence of robot motion has been obtained through nonlinear constraint optimization, which takes into consideration the visibility loss of features due to camera motion. Using a visual servoing simulator, the predictive strategy was successfully tested.