Autonomous landing on the deck of a boat or an unmanned surface vehicle (USV) is the minimum requirement for increasing the autonomy of water monitoring missions. This paper introduces an end-to-end control technique based on deep reinforcement learning for landing an unmanned aerial vehicle on a visual marker located on the deck of a USV. The solution proposed consists of a hierarchy of Deep Q-Networks (DQNs) used as high-level navigation policies that address the two phases of the flight: the marker detection and the descending manoeuvre. Few technical improvements have been proposed to stabilize the learning process, such as the combination of vanilla and double DQNs, and a partitioned buffer replay. Simulated studies proved the robustness of the proposed algorithm against different perturbations acting on the marine vessel. The performances obtained are comparable with a state-of-the-art method based on template matching.

This paper presents a collision-free motion planning method of two articulated robot arms in a three dimensional common work space. Each link of a robot arm is modeled by a cylinder ended by two hemispheres, and the remaining wrist and hand is modeled by a sphere. To describe the danger of collision between two modeled objects, minimum distance functions, which are defined by the Euclidean norm, are used. These minimum distance functions are used to describe the constraints that guarantee no collision between two robot arms. The collision-free motion planning problem is formulated as a pointwise constrained nonlinear minimization problem, and solved by a conjugate gradient method with barrier functions. To improve the minimization process, a simple grid technique is incorporated. Finally, a simulation study is presented to show the significance of the proposed method.

This paper considers planning and control algorithms that enable a programmable sheet to realize different shapes by autonomous folding. Prior work on self-reconfiguring machines has considered modular systems in which independent units coordinate with their neighbors to realize a desired shape. A key limitation in these prior systems is the typically many operations to make and break connections with neighbors, which lead to brittle performance. We seek to mitigate these difficulties through the unique concept of self-folding origami with a universal fixed set of hinges. This approach exploits a single sheet composed of interconnected triangular sections. The sheet is able to fold into a set of predetermined shapes using embedded actuation.

We describe the planning algorithms underlying these self-folding sheets, forming a new family of reconfigurable robots that fold themselves into origami by actuating edges to fold by desired angles at desired times. Given a flat sheet, the set of hinges, and a desired folded state for the sheet, the algorithms (1) plan a continuous folding motion into the desired state, (2) discretize this motion into a practicable sequence of phases, (3) overlay these patterns and factor the steps into a minimum set of groups, and (4) automatically plan the location of actuators and threads on the sheet for implementing the shape-formation control.

Part 1 of the paper* was concerned with the Jacobian of serial robot-arms and its matrix of cofactors; part 2 emphasizes the end-effector's velocity characteristics as affected at special configurations. The displacement about a ‘fixture-point’ is extended dually to a ‘fixture-plane’, and then to a ‘fixture-line’. These fixture-elements serve to highlight the prolixity of special configurations. Reciprocal screws lead to a complete survey of how the end-effector can ‘twist’ when one freedom is lost; extension to loss of more freedoms is explained. Then the fixture-elements are given elementary displacements that allow common requirements to be studied with the minimum of complications. Further investigation is thought to have promise for robot-control.

The intention of this paper is to contribute towards a unified understanding of the basic notions and terms in the domain of humanoid robotics, having in mind that the same notions are sometimes interpreted in different ways (some interpretations are contradictory, and some even erroneous). Hence, the first part of the paper is devoted to defining some basic notions, walk and gait being among the first. Then, the paper deals with the notion of dynamic balance and stability, particularly the difference between them, since these essentially different notions are often confused and, rarely, regarded as identical. As dynamic balance is directly related to the notion of zero-moment point (ZMP), it was necessary to touch upon some misunderstandings concerning the ZMP. Gait stability is an especially delicate category, as humanoid locomotion systems have certain specific features that are not possessed by other systems. Namely, because of external disturbances, there may appear unpowered (passive) degrees of freedom that cause loss of dynamic balance. Hence, these unpowered degrees of freedom cannot be overlooked in the stability analysis. As the stability of motion of humanoid robots is inseparably linked with control, it was also necessary to pay due attention to this notion. Finally, the paper ends with a discussion of posture and postural stability with all their specificities. The authors hope that this paper will contribute to a clearer understanding of the basic notions of humanoid robotics, especially concerning robots with high dynamic and control performances.

In this paper, an efficient motion planning method is proposed for a six-legged robot walking on irregular terrain. The method provides the robot with fast-generated free-gait motions to traverse the terrain with medium irregularities. We first of all introduce our six-legged robot with legs in parallel mechanism. After that, we decompose the motion planning problem into two main steps: first is the foothold selection based on a local footstep cost map, in which both terrain features and the robot mobility are considered; second is a whole-body configuration planner which casts the problem into a general convex optimization problem. Such decomposition reduces the complexity of the motion planning problem. Along with the two-step planner, discussions are also given in terms of the robot-environmental relationship, convexity of constraints and robot rotation integration. Both simulations and experiments are carried out on typical irregular terrains. The results demonstrate effectiveness of the planning method.

In this paper, the authors use the conformal geometric algebra in robotics. This paper computes the inverse kinematics of a robot arm and the differential kinematics of a pan–tilt unit using a language of spheres showing how we can simplify the complexity of the computations.

This work introduces a new geometric Jacobian in terms of bivectors, which is by far more effective in its representation as the standard Jacobian because its derivation is done in terms of the projections of the involved points onto the line axes. Furthermore, unlike the standard formulation, our Jacobian can be used for any kind of robot joints.

In this framework, we deal with various tasks of three-dimensional (3D) object manipulation, which is assisted by stereo-vision. All these computations are carried out using real images captured by a robot binocular head, and the manipulation is done by a five degree of freedom (DOF) robot arm mounted on a mobile robot. In addition to this, we show a very interesting application of the geometric Jacobian for differential control of the binocular head. We strongly believe that the framework of conformal geometric algebra can generally be of great advantage for visually guided robotics.

This paper presents a method for designing the arm lengths of a screw drive in-pipe robot with a pathway selection mechanism to pass through bent pipes. The robot comprises a front-rotating unit (rotator), middle-steering unit, and rear-supporting unit (stator). Each unit has elastic arms, which can be elongated by springs and can be pushed against the inner wall of the pipe. The robot can travel and steer not only in straight pipes but also in bent pipes and T-branches. When the robot passes through such pipes, its mobility depends on whether the arms can touch the inner wall because the robot will lose the driving force if one of the arms cannot reach the inner wall. This is especially important in vertically positioned pipes, where the robot will fall down if it is unable to reach. In this paper, we focus on the cross-section of the bent pipes, arm length, spring stiffness, and mass of the robot to clarify the mobility of the robot in these types of pipes. A method for determining the initial arm length on the basis of simple quasi-statics is proposed and experimental verification is conducted.

Surgical and search/rescue robots often work in environments with very strict spatial constraints. The tendon-sheath mechanism is a promising candidate for driving such systems, allowing power sources and actuation motors placed outside to transmit force and energy to the robot at the distal end through the constrained environment. Having both compactness and high force capability makes it very attractive for manipulation devices. On the other hand, the friction attenuation of tendon tension is nonlinear and configuration-dependent due to tendon/sheath interactions throughout the transmission path. This is a major obstacle for the tendon-sheath mechanism to be widely adopted. Here, we focus on the friction analysis for flexible and time-varying tendon-sheath configurations: the most challenging but yet commonly encountered case for real-world applications. Existing results on fixed-path configurations are reviewed, revisited, and extended to flexible and time-varying cases. The effect of tendon length to friction attenuation is modeled. While focusing on tension transmission, tendon elongation is also discussed with the length effect applied. In the end, two-dimensional results are extended to three-dimensional tendon-sheath configurations. All propositions and theorems are validated on a dedicated experimental platform.

This paper reviews the state-of-the-art in robotic tails intended for inertial adjustment applications on-board mobile robots. Inspired by biological tails observed in nature, robotic tails provide a separate means to enhance stabilization, and maneuverability from the mobile robot's main form of locomotion, such as legs or wheels. Research over the past decade has primarily focused on implementing single-body rigid pendulum-like tail mechanisms to demonstrate inertial adjustment capabilities on-board walking, jumping and wheeled mobile robots. Recently, there have been increased efforts aimed at leveraging the benefits of both articulated and continuum tail mechanism designs to enhance inertial adjustment capabilities and further emulate the structure and functionalities of tail usage found in nature. This paper discusses relevant research in design, modeling, analysis and implementation of robotic tails onto mobile robots, and highlight how this work is being used to build robotic systems with enhanced performance capabilities. The goal of this article is to outline progress and identify key challenges that lay ahead.

Motion coupling is one of the typical characters of parallel robots. This makes it complicated to control the parallel robots. Therefore, the design of parallel mechanisms with decoupled motions is an important and challenging issue in the parallel robotic field. However, research in this field, including the definition and type synthesis theory, is weak or lacking. In this paper, the reducible correlation between the input and the output of 3-DOF translational mechanisms with decoupled motions is investigated and the definition of this topic is presented. A new parameter named Gf coordinate is propsed for the synthesis of 3-DOF reducible translational mechanisms. The type synthesis theory for this kind of mechanisms is obtained and some reducible mechanisms are synthesized.

In this paper, two novel approaches to unmanned underwater vehicle path planning are presented. The main idea of the first approach, referred to as Constrained Optimisation (CO) is to represent the free space of the workspace as a set of inequality constraints using vehicle configuration variables. The second approach converts robot path planning into a Semi-infinite Constrained Optimisation (SCO) problem. The function interpolation technique is adopted to satisfy the start and goal configuration requirements. Mathematical foundations for Constructive Solid Geometry (CSG), Boolean operations and approximation techniques are also presented to reduce the number of constraints, and to avoid local minima. The advantages of these approaches are that the mature techniques developed in optimisation theory which guarantee convergence, efficiency and numerical robustness can be directly applied to the robot path planning problem. Simulation results have been presented.

In this paper, methods based on various spline techniques for planning and fast modifications of a trajectory for robot manipulators are investigated. Algebraic and trigonometric splines, their combined use, and the use of the B-spline technique are analyzed and compared in detail. In so doing, we focus on the performance of sudden changes in a predefined trajectory, e.g. obstacle avoidance in real-time applications. Some comparative examples illustrate our results.

For the trajectory following problem of a robot manipulator, a robust estimation and control scheme which requires only position measurements is proposed to guarantee uniform ultimate bounded stability under significant uncertainties and disturbances in the robot dynamics. The scheme combines a class of robust control laws with a robust estimator where the robust control law can be chosen to be either a modification of the standard computed torque control law or simply a linear and decentralized “PD” control law. The proposed robust estimator is also linear and decentralized for easy implementation. Constructive choices of the gains in the control law and estimator are proposed which depend only on the coefficients of a polynomial bounding function of the unknown dynamics. The asymptotic stability of the tracking errors and the estimation error is also investigated. Experimentation results verify the theoretical analysis.

Complex tasks that need to be performed through teleoperation led to the development of multifinger robot hands. A dextrous master is a multi-DOF controller which is worn by the operator in order to teleoperate these anthropomorphic hands. Force feedback is very useful when there is interaction with the environment. Providing force feedback to dextrous masters is an area of active research. We outline some of the human factors that influence the design of such masters. Of obvious importance is the hand geometry and the desired number of degrees of freedom. Additional criteria relate to the force perception of the hand. Finally, the man-machine impedance is of importance, since at the man/machine interface there are two impedances acting in series. One is the effective impedance of the human operator holding the master controller, the second is the impedance of the master controller being manipulated.

Using simple running models, researchers have argued that swing-leg retraction can improve running robot performance. In this paper, we investigate whether this holds for a more realistic simulation model validated against a physical running robot. We find that swing-leg retraction can improve stability and disturbance rejection. Alternatively, swing-leg retraction can simultaneously reduce touchdown forces, slipping likelihood, and impact energy losses. Surprisingly, swing-leg retraction barely affected net energetic efficiency. The retraction rates at which these effects are the greatest are strongly model-dependent, suggesting that robot designers cannot always rely on simplified models to accurately predict such complex behaviors.

The effect on failure rates of the way tele-operators interact with mobile robots is investigated. Human tele-operators attempted to move a robot through progressively more complicated environments with reducing gaps, as quickly as possible. Tele-operators used a joystick and either watched robots, while operating them, or used a computer screen to view scenes remotely. Cameras were either mounted on the robot to view the space ahead of the robot or mounted remotely so that they viewed both the environment and robot. Tele-operators completed tests both with and without sensors. Both an umbilical cable and a radio link were used.

Optical tweezers have been used as versatile tools for non-contact manipulation of micrometer-sized entities. This paper proposes a hybrid micro/nanoscale manufacturing system using optical tweezers and chemical linkages for fabricating 2D and 3D micro/nanostructures. A holographic multiple trap optical tweezers system is first used to trap particles in a desired pattern. The particles are then connected to form rigid units using suitable chemistry. Connection schemes based on gold seeding, complementary-DNA linkage and streptavidin-biotin chemistry are presented and possible applications of this technique are explored. This method combines the advantages of top-down and bottom-up approaches and is compatible with organic and inorganic materials.

A method is presented for the type synthesis of a class of parallel mechanisms having one-dimensional (1D) rotation based on the theory of Generalized Function sets (GF sets for short), which contain two classes. The type synthesis of parallel mechanisms having the first class GF sets and 1D rotation is investigated. The Law of one-dimensional rotation is given, which lays the theoretical foundation for the intersection operations of GF sets. Then the kinematic limbs with specific characteristics are designed according to the 2D and 3D axis movement theorems. Finally, several synthesized parallel mechanisms have been sketched to show the effectiveness of the proposed methodology.