ROMANSY 22 – Robot Design, Dynamics and Control

In this keynote paper challenges in Robotics are discussed in terms of Innovation issues coming from Mechanism Design as they were and still are fundamental for achieving developments and technological transfer from Mechanism and Machine Science into novel successful mechanical designs of modern robot systems.

Robotics is undergoing a major transformation in scope and dimension with accelerating impact on the economy, production, and culture of our global society. The generations of robots now being developed will increasingly touch people and their lives. They will explore, work, and interact with humans in their homes, workplaces, in new production systems, and in challenging field domains. The emerging robots will provide increased support in mining, underwater, hostile environments, as well as in domestic, health, industry, and service applications. Combining the experience and cognitive abilities of the human with the strength, dependability, reach, and endurance of robots will fuel a wide range of new robotic applications. The discussion focuses on design concepts, control architectures, task primitives and strategies that bring human modeling and skill understanding to the development of this new generation of collaborative robots.

Partitioned approximation control is avoided in most decentralized control algorithms; however, it is essential to design feedforward control with improved tracking accuracy. As a result, this work is focused on decentralized adaptive partitioned approximation control for complex robotic systems using the orthogonal basis functions as strong approximators. In essence, the partitioned approximation technique is intrinsically decentralized with some modifications. The proposed decentralized control law consists of three terms: the partitioned approximation-based feedforward term that is necessary for precise tracking, the high gain-based feedback term, and the adaptive sliding gain-based term for compensation of modeling error. The passivity property is essential to prove the stability of local stability of the individual subsystem with guaranteed global stability. A two-link robot is simulated to verify the effectiveness of the proposed technique.

A novel robotic leg mechanism based on a parallel architecture is described in this paper. A walking operation of the mechanism is defined and characterized as function of its main design parameters through a numerical simulation. The same operation is then performed by a prototype of the proposed design that has been manufactured through 3D printing. Finally, the results of the numerical simulation and of the experimental tests are compared and discussed.

Flexible robots can be modeled as underactuated multibody systems since they generally have less control inputs than degrees of freedom for rigid body motion and deformation. The flexibilities must be taken into account in the control design. In order to obtain high performance in the end-effector trajectory tracking, an accurate and efficient nonlinear controller is required. In this paper, a nonlinear feedback controller based on the feedback linearization approach using all the states of the system is designed and carefully tested on a very flexible parallel lambda robot. The simulation and experimental results show that the end-effector tracks a trajectory with higher accuracy compared to previous works.

The subject of this paper is the elastostatics of a novel three-limb, full-mobility parallel-kinematics machine (PKM) with flexible links, dubbed the SDelta robot. Due to the inherent flexibility of the light-weight limb rods under fast operations, they are modeled as identical linearly elastic beams. The moving platform of the PKM is assumed to be elastically attached to the base platform via a six-dof generalized spring. Because of the symmetric design of the SDelta robot, the constant generalized stiffness matrix becomes isotropic. Moreover, the posture-dependent Cartesian stiffness matrix is derived. Based on the preliminary design of a prototype at the desktop scale, its stiffness matrix is numerically evaluated.

Piping inspection robots are of great importance in industries such as nuclear, sewage and chemical where the internal diameters of the pipeline are significantly smaller. Mechanisms having closed loops can be used in such areas as they generate contact forces and deployable structures. With the help of a bio-inspired mechanism, a piping inspection robot is presented which mimics the motion of a caterpillar. The robot is composed of three modules: a central module for elongation and two other modules on the front and rear for clamping. A slot-slider mechanism is chosen for the legs of the robot. Using industrial components such as DC motors, servo-controllers, ball screws and fasteners, the entire robotic system was realized in CATIA software and a prototype was made at the Laboratoire des Sciences du Numérique de Nantes (LS2N). In this article, we present the forces induced on the motors under locomotion using a dynamic analysis. With the help of the recursive Newton-Euler algorithm, the torques generated on the motor under locomotion have been identified which ensures the stability of the system while moving inside pipes.

The paper presents the structural design and motion simulation of the robotic eyes with the characteristics of the female eye. The eyeballs/eyelids drive system with total 7 DOFs is proposed. The eyeballs drive system has 3 DOFs and enables the rotation of both eyeballs together around the pitch axis and their independent rotation around the yaw axis. The proposed solution enables the installation of cameras directly into the eyeballs. Also, it is possible to independently rotate the upper/lower eyelids of each eye, where the eyelids, in relation to the eyeballs, are rotated in two planes. Motion simulation of eyeball and upper/lower eyelids is performed. For the range of the motion that corresponds to the human eye, the angular velocities of the eyeball and upper/lower eyelids reach and exceed the kinematic parameters of the human eye.

Recently, variable stiffness actuators (VSAs) have been introduced for reducing the input efforts of pick-and-place robots. However, the serial arrangement of springs and motors in the VSAs decreases the accuracy at high-speeds due to uncontrolled robot deflections. To ensure accuracy while reducing the input efforts, this paper proposes the use of variable stiffness springs (VSS) in parallel configuration with the motors. The parallel arrangement of VSS and motors is combined with a shooting method to adjust the stiffness of the system in order to enforce its limit cycle to converge to a desired trajectory, and, thus, to decrease the input torques. Numerical simulations of the suggested approach on a five-bar mechanism show the reduction of the robot input efforts.

The sub-optimality of redundant manipulators inverse kinematics solutions deriving from calculus of variations is discussed through some considerations of topology. This paper clarifies the relations of homotopy linking together, on one hand, self-motion manifolds and, on the other, joint space paths. With an example, it proves that the sub-optimality does not depend on the homotopy classes of self-motions but is linked to both the presence of distinct self-motions and the existence of different homotopy classes of joint space paths. This paper also clarifies the notion of pre-image of a workspace path, showing that more complex surfaces than deformed tori can be generated in the Cartesian configuration space depending on some topological properties of the path.

Redundancy resolution schemes based on calculus of variations present several drawbacks limiting the intrinsic potential, in terms of augmented dexterity and flexibility, of redundant manipulators. In particular, they do not guarantee the achievement of the globally-optimal solution. Grid search algorithms can be designed starting from dynamic programming (DP) which overcome the limits of calculus of variations. This paper, in particular, presents a novel algorithm that considers the employment of multiple DP grids to be searched together at the same time. Such a technique achieves the global optimum, while allowing for pose reconfiguration of the manipulator while the task is executed.

While running, humans use the stiffness of the knee and ankle joint of the leg. Mimicking this motion can improve the output power and performance of humanoid robots. It also offers the possibility of clarifying running in humans, from an engineering perspective, by mimicking other characteristics of an ankle joint. In this paper, we design an ankle and foot mechanism that mimics human’s characteristics, such as joint stiffness in the direction of pitch, and following the floor surface in the direction of roll upon landing for stabilization. To mimic these characteristics, our ankle joint mechanism consisted of CFRP (Carbon Fiber Reinforced Plastic)-laminated leaf springs implemented on the foot of the robot for a deflection in direction of pitch and a twist in the direction of the roll. We ensured that the ankle joint can follow the ground in the direction of roll at landing in a hopping experiment.

This paper presents a cable-driven assisting device, which has been designed and built at LARM in Cassino. Experimental tests are presented as the basis for design improvements aiming to achieve a portable user-oriented solution of the cable-driven assisting device for applications in human upper limb exercising and rehabilitation. Experimental tests are reported to show the engineering feasibility and soundness of the proposed solutions.

The mobility of wheeled or orbital mobile welding robot is constrained by obstacles, uneven terrain or orbit. A new legged mobile welding robot which can surmount obstacles and move freely on uneven terrain is presented in this paper. This robot is the combination of a hexapod robot and a six DOF manipulator. Each leg of the hexapod robot is a three DOF parallel mechanism which has excellent load bearing capability. The manipulator is a PRPRRR mechanism which is characterized by motion decoupling. Inverse kinematic model of this robot is built, and workspace analysis is presented. Then, motion simulation of a typical welding workflow is delineated.

The paper introduces new methods for EMG signals analysis. Two types of artificial neural networks have been applied for this purpose, namely the learning vector quantization (LVQ) neural network and the competitive neural network. The sets of EMG signals were recorded for different walking conditions. The gait of a healthy person without muscle disorders history were recorded. The study proved that the LVQ neural network provides a good tool for detecting the differences in the EMG trajectories coherent with the different effort of the muscles work, and the competitive neural network performs well the EMG clustering. The last property allows to identify the motion fragments with similar muscles effort and to predict the farther continuation of movements after recording the initial fragments of EMG. Obtained results are useful for robotic applications and could use for the reference for exoskeleton design and rehabilitation purposes.

The existing mode for exploratory mission of extraterrestrial body has some limitations of the exploratory range. In order to widen the exploratory range, a lander is expected to walk, i.e., it will be a robot with the features of lander and rover. Furthermore, one of the most challenges to design the robot is to deal with the contradiction of degree of freedom (DoF) during different phase: during landing phase, the lander is regarded as a truss or structure without motion, while during walking phase, the lander is regarded as a mechanism with particular motions. In this paper, a general approach of type synthesis of mechanisms generated from a given truss is investigated. Adopting this approach, numerous structures of legged mobile landers are obtained.

A vibration-driven locomotor (capsubot) consisting of a rigid housing and an internal body connected to the housing by a spring is considered. The system is driven and controlled by an electromagnetic actuator that provides a force interaction between the housing and the internal body. The housing moves along a line on a horizontal plane with dry friction. The control voltage is applied to the robot in a periodic pulse-width mode, the voltage polarity remaining unchanged. Theoretical analysis predicts that the speed and direction of motion of the robot can be controlled by varying the period or/and the duty cycle of the control signal. An experimental prototype of the robot is built and the experiments are performed. The experiments confirm the theoretical prediction.

In this paper, harmonic drive which is commonly used in robot joints is studied. A kinematic model of harmonic drive with input eccentricity error is established. The model and the deformation compatibility equation are established to solve the motion trajectory of a point on the neutral layer, the backlash distribution and the load distribution. Through an example analysis, it is found that the input eccentricity error makes the motion law of each point on the neutral layer no longer the same, and makes the backlash distribution uneven and the load concentrated in the eccentric direction. The model established in this paper can also provide a basis for the establishment of assembly tolerance.

The proposed research presents a novel kinematic design of rotary hexapod actuated by single drive. The developed mechanism has been designed as a combination of spatial kinematic chain 6SS and planar gear lever linkage set inside of a circular guide and serves as a movable base for hexapod’s legs. The planar mechanism includes six kinematic chains ending with carriages that set under each leg of spatial 6SS chain. Motions of the hexapod’s platform are predefined and controlled by single rotational drive set in the center of a circular guide that is mounted on a fixed link. Novel design of the proposed rotary hexapod allows changing trajectories of its platform by variation of only one link length in each kinematic chain of the movable base. The hexapod can be used as a system for spatial orientation of various objects, applied in car or airplane simulators as well as in rehabilitation equipment that requires cyclic movements of an end-effector.

In this paper, the dynamic stability problem of a parallel wrist mechanism is studied by means of monodromy matrix method. This manipulator adopts a universal joint as the ball-socket mechanism to support the mobile platform and to transmit the motion/torque between the input shaft and the end-effector. The linearized equations of motion of the mechanical system are established to analyze its stability according to the Floquet theory. The unstable regions are presented graphically in various parametric charts.

This paper analyzes some parallelisms between 3UPS-PU Tricept-like parallel robots and their planar version, 2RPR-PR manipulators, in terms of forward kinematics and singularities. We show that, like 2RPR-PR manipulators, all 3UPS-PU robots with flat mobile platform have special singularities for which the mobile and fixed platforms are coplanar. These special singularities turn out to be eightfold solutions of the forward kinematics, their perturbations result in double deltoids and, unlike in 2RPR-PR manipulators, encircling them does not produce nonsingular transitions.

Gravity balancing is used to reduce the actuator effort in robotic systems. However, it is well known that conventional linear springs, directly jointed with a rotating link cannot ensure a complete gravity balancing. To achieve a complete balancing of a rotating link the zero length spring should be applied. The zero length spring corresponds to a spring with special coils ensuring zero elastic force for zero length of the spring. Therefore, the use of traditional springs, i.e. non-zero length springs, leads to adding of auxiliary mechanisms. The aim of the present study is to propose an analytically tractable solution permitting to synthesize a planar four-bar linkage that will provide a more optimal generation of the balancing moment developed by a non-zero length spring. The efficiency of the suggested linkage design is illustrated via numerical simulations. It is shown that a quasi-perfect balancing has been achieved.

Depending on the modeling approach, dynamic analyses of parallel robots often involve expensive computations of undesired constraint forces, cumbersome partial derivatives, or large matrix operations. Decomposing the parallel structure into virtual tree structures and the free platform allows for reducing the system size as well as efficient recursive and parallel computing. Against this background, the Khalil-Ibrahim Method is adapted to the well-known Delta parallel robot. It is shown that the resulting model deftly removes constraint forces while being modularly expandable for analyzing recent design modifications using additional serial chains attached to the parallel Delta structure.

The aim of this paper is to propose a geometric based approach for workspace analysis of Translational Parallel Manipulators (TPMs), which will be useful for the dimensional synthesis problem. For this purpose, a non-exhaustive list of TPMs in the bibliography is presented and grouped according to the structure of their legs as well as the shape of their workspaces. The approach is easy to implement and it is illustrated through three TMPs examples, having each a different workspace shape.

A mathematical model of a wind car is introduced. The car is equipped by the Magnus-type horizontal axis wind rotor. The shaft of the rotor is connected by a reduction gear with driving wheels of a vehicle. The wheels move without slipping along the line that is parallel to the wind flow. Steady motions of the vehicle are studied. It is shown that the vehicle can move upwind, as well as downwind faster than the wind. The speed of the vehicle is estimated depending on the transmission gear ratio. The maximal speed is compared with that obtained for different types of wind powered vehicles.

This paper deals with the interaction between a specified robot and its environment. A particular case is considered, which is a co-manipulation case, where an operator is performing a co-operative task with the robot. Cascaded loops are considered for the control design and a frequency analysis is performed to study the influence of the position loop.

This paper presents the static equilibrium workspace of an under-constrained cable-driven robot with four cables taking into account the forces and the moments due to the forces acting on the moving platform. The problem is formulated as a non-linear optimization problem with maintaining static equilibrium as the objective function. The simulations are done using MATLAB. The maximum force on the cables and tilting angle of the platform are used to define the feasible static equilibrium workspace and the results obtained are used to finalize the design of the collaborative cable-driven robot to be installed in existing production lines for the agile handling of parts in a manufacturing industry.

The paper describes the design and application of a low-cost haptic device with five degrees of freedom (5-DOF), all of which are fully actuated for force/torque haptic display. The device is realized using two Novint Falcons, commercially available haptic devices with only translational DOF. The paper focus on the practical issues related to the use of such a haptic device as a robotic teleoperation master and provides solutions to improve its usability during an application inspired by surgical robotics, namely the teleoperated insertion of a needle into a soft tissue. Experiments demonstrate the usefulness of force feedback and virtual fixtures on both translation and orientation degrees of freedom.

The operational workspace of parallel robots is often reduced by the presence of singularities. Recently, it has been proven that Type 2 singularities can be crossed in a way such that the dynamic model of the robot never degenerates. This discovery has been the starting point of several works on multi-model Computed Torque Control (CTC) that allow crossing of Type 2 singularities. In this paper, we propose a further improvement thanks to adaptive control. The major contribution of the paper is in the control law synthesis, which uses only linear methods, in contrast to usual approaches based on Lyapunov theory. This theoretical development will be validated both in simulation and experimentally.

The aim of this paper is to define a weighting function to achieve a coordinated control of both parts of a mobile manipulator. The function presented here depends on the desired location of the end-effector with respect to the workspace of the robotic arm. The objective is to optimize the workspace of the manipulator, such that phenotyping can be executed without further motion of the mobile base. Experimental results are presented, carried out on a system composed of a 6-dof robotic arm and a gantry.

This paper deals with a model-based feed-forward torque control strategy of non-redundant cable-driven parallel robots (CDPRs). The proposed feed-forward controller is derived from an inverse elasto-dynamic model of the CDPR to compensate for the dynamic and oscillatory effects due to cable elasticity. A PID feedback controller ensures stability and disturbance rejection. Simulations confirm that tracking errors can be reduced by the proposed strategy compared to conventional rigid body model-based control.

This paper describes a path planning framework for processing 2D maps of given pedestrian locations to provide sidewalk paths and street crossing information. The intention is to allow mobile robot platforms to navigate in pedestrian environments without previous knowledge and using only the current location and destination as inputs. Depending on location, current path planning solutions on known 2D maps (e.g. Google Maps and OpenStreetMaps) from both research and industry do not always provide explicit information on sidewalk paths and street crossings, which is a common problem in suburban/rural areas. The framework’s goal is to provide path planning by means of visual inference on 2D map images and search queries through downloadable map data. The results have shown both success and challenges in estimating viable city block paths and street crossings.

This study proposes “Human-Powered Robotics,” a concept of the intelligent machines driven by the power applied by user directly. The development of the robotic personal mobility vehicle prototype based on this concept is described in this paper. This is a vehicle driven by the operator’s pedaling power, which is capable of servo control of the wheel angle using the powder clutch. Experimental results show that it is possible to perform servo control of the wheels and traveling control including forward/backward/pivot turn, and to stabilize the posture as an inverted pendulum by the operator’s pedaling power.

Robotic Friction Stir Welding (RFSW) is an innovative process which allows the joining of aluminum materials with robots. The main drawback of using robots to perform friction stir welding is that the end effector of the robots deviates in position and orientation during welding. This is due to the high forces induced by the process and the weak stiffness of the robots. Thus, the compensation of the deviations must be taken into account in an RFSW path planning. In this paper, a methodology based on B-splines curves is proposed to generate offline a welding path with the compensation of the elastic deformation of a robot in order to achieve a high quality welding of three-dimensional workpieces. The methodology is validated on a Kuka robot KR500-2MT which performs a sinusoidal welding path defined on a cylindrical surface. The experiment shows the effectiveness of the methodology and allows to reduce significantly (about $$88\%$$88%) the lateral end effector deviation.

This paper proposes an iterative learning scheme for in-hand manipulation systems by utilizing the learning gain adaptation concept of deep learning. The advantages of the proposed method are that (1) there is no need to generate theoretical analytical models for the learning process and (2) the proposed method is robust against uncertainties such as measurement errors, friction force, and contact state. Finally, the validity of the proposed method is verified through experiments.

In this paper, we propose a body mechanism for a legged robot to reduce slippage when moving on an uneven terrain. In the proposed landing mechanism, linear spikes are distributed on the grounding, and they move passively with respect to the road surface shape. This mechanism is aimed to improve the grip performance on an irregular road surface and the movement capability. The developed mechanism is mounted on the body of a four-limbed robot named “WAREC-1.” The robot performs crawling motion over an inclined rough terrain. The crawling includes a motion wherein its legs and torso alternately contact the ground. The experimental results show that the downward sliding of the robot is reduced and that the mechanism contributes to improving the grip performance and movement capability on uneven terrain.

In this paper, we present a simulator that has been developed using Gazebo and ROS to study cable-driven parallel robots. Real-time dynamic simulation of such robots is an efficient approach to develop new control laws that may integrate various sensors. The limitations of Gazebo are dealt with, as we model the cables under tension as massless $$U-\underline{P}-S$$U-P̲-S links with the prismatic joint actuated. We illustrate the proposed simulator with a dynamic controller, detailing the tension distribution and performing various trajectories.

Constructive scheme for the design of a multi-purpose portable movable orthosis with elastic elements is proposed. The static balancing cases of biomechanical systems with this orthosis-assistant are considered during waking and sitting. This orthosis can be applied for the physiotherapy, movement support, exercises of human walking and sit-to-stand. A numerical analysis of the scheme is carried out; its operating modes, advantages and disadvantages are revealed.

This paper presents a preliminary survey of the use of direct drive linear motors for joint actuation of a humanoid robot. Their prime asset relies on backdrivability, a significant feature to properly cushion high impacts between feet and ground during dynamic walking or running. Our long-term goal is the design of high performance human size bipedal walking robots. However, this paper focuses on a preliminary feasibility study: the design and experimentation of a mono-actuator lower limb.

Humanoid robots should prevent damage during falling forwards, which will improve the ability of serving people on various occasions. In this paper, we analyze different landing positions, and propose a falling forwards protection strategy for humanoid robots. Falling forwards is divided into two phases: the knee landing phase and the chest landing phase. The parametric optimal strategy based on telescopic inverted pendulum with flywheel is used to plan the motion of robot in the first phase to reduce the impact. Inverted pendulum with flywheel and fixed length is adopted in the second phase. The simulations tested on the BHR-6P (BIT Humanoid Robot 6 Prototype) platform validate the effectiveness of the presented strategy.

In the field of robotics, small cat-like robots, due to its small size, good flexibility, low energy consumption, has become a research trend. Elastic four-bar linkage mechanism (EFLM) with programmable trajectory, good stability, and certain buffer performance, is an excellent choice for driver design of small quadruped robot. We designed a novel miniature cat-like robot, Og-cat, which is optimized in structural features and special parameters compared with other similar quadruped robots. Moreover, we developed a method that can make leg trajectory implement accurately for robots using EFLM. A trajectory realization function is obtained through MATLAB fitting so that the trajectory can be executed precisely. The proposed method has been confirmed on our small quadruped robot Og-cat.

Eco-design of robots is a field of research which has been rarely explored in the past. In order to considerably decrease the environmental impact of robot during the design phase, metal or carbon composite parts can be replaced by bio-sourced materials, such as wood. Indeed, wood has interesting mechanical properties, but its performance/dimensions will vary with the atmospheric conditions/external solicitations and with the conditions in which trees have grown. This paper deals with the design of a stiff and accurate wooden five-bar mechanism. First, a control-based design problem is formulated. This problem aims at finding the optimal parameters of the robot, taking into account the nature of the desired control (sensor-based control). Then, reliable topology optimization methodology is proposed, taking into account the variability of the wood performance and that will allow the definition of robot architectures (shape of the links) for which the impact of this variability in terms of deformation is minimal. Finally, the optimal design variables are given and are used for the realisation of industrial prototype of a wooden five-bar mechanism.

The paper deals with modeling of a thermomechanical actuator for a microrobot. The actuator is modeled by a linkage of rigid bars connected by elastic joints. A series of experiments were performed to estimate the parameters of the adopted model. The parameters of the model are calculated and validated. The actuators of such a type can be used, for example, in robotic devices designed for operation in spacecraft. These actuators are simple to manufacture and control. Mathematical modeling of the mechanical properties of the actuators could help plan motions and design control strategies for robotic systems.

In this paper a stiffness analysis approach for Delta parallel robots is presented, which combines the Virtual Joint Method (VJM) with a Finite Element Analysis (FEA) stiffness model of the proximal link. By comparison with a purely analytical model and measurements it is shown that using this more laborious combination of VJM and FEA is reasonable in early stages of the robot design process.

The paper presents advancement of the matrix structural analysis technique (MSA) for stiffness modeling of robotic manipulators. In contrast to the classical MSA, it can be applied to both parallel and serial manipulators composed of flexible and rigid links connected by rigid, passive or elastic joints with multiple external loadings. The manipulator stiffness model is presented as a set of basic equations describing the link elasticities that are supplemented by a set of constraints describing connections between links. These equations are aggregated straightforwardly in a common linear system without traditional merging of the matrix rows and columns, which allows avoiding conventional manual transformations at the expense of numerical inversion of the sparse matrix of higher dimension.

This paper presents an attempt to robotize the grinding process and to overcome grinding vibrations and chattering. The objective is to have a finished workpiece with a high quality of the final surface. In order to achieve that, we started by choosing the right strategy to grind the workpiece that has uneven initial surface. Then, a well-known model of the process is used in order to simulate the grinding of a metallic workpiece. The robot is supposed rigid and does not contribute in the flexibility of the system. The only flexibility that was taken into consideration is that of a pneumatic actuator used to control and reduce vibrations. Its dynamic behavior is approximated using a second degree transfer function.

This paper proposes a mathematical model of the walking with canes for an anthropomorphic biped with two identical legs with massless feet, two identical arms, and a torso. The walking is performed in the sagittal plane. The period of the walking gait is the stride that is composed of single support (SS) phase on the first leg, an instantaneous double support (DS) phase, a SS phase on the other leg and another instantaneous DS phase. The stride is the period of this cyclic walking, because the motion of coupled arms is synchronized on this stride. The cane is considered massless. Thus, in order to compare the walking with and without massless cane, the same dynamic model is considered. Numerical tests show that the magnitude of the ground reaction in the stance foot is less with a massless-cane assistance than without one. Especially, the results prove that it is better to use canes with a handle that allows to apply on it a force and a moment by the user. These results highlight the importance of handles in the designing process of the canes. This theoretical study may benefit the design of new canes to overcome a disability on the lower limbs with further researches.

This paper presents a synthesis method for the Autonomous Underwater Vehicle (AUV) propulsion system based on the features of a trajectory to follow. This method is based on solid/fluid dynamics analysis of a AUV performing the trajectory following task and gives actuation requirements to achieve it properly. This actuation is then used to generate a propulsion system under the form of a Thrust Configuration Matrix (TCM) that is compatible with the desired trajectory. From this matrix the number of thrusters, their position and direction can then be extracted to synthesis a fitted propulsion system. Thus, the propulsion capabilities of the robot will match the trajectory characteristics and it will be able to follow it. The objective of this work is to provide an evaluation as well as a design method to produce a controllable AUV for a given task. A second use of such an analysis is to provide an evaluation process allowing to perform AUV task-based design. The method is developed for generic fixed-thrusters AUVs on any trajectory, and is applied to an existing 4-thrusters AUV for a seabed scanning flat trajectory. The method shows why the initial AUV propulsion does not match the task and what is the solution solving the issue.

In this paper, a musical articulation system for the Waseda Anthropomorphic Saxophonist robot is proposed. First, the specifications to produce musical articulation with a saxophone are determined. Then, a tonguing mechanism and a fast air flow control valve are implemented on the robot along with an integrated feedback control system to perform articulation by synchronously stopping reed vibration and reducing the air flow. The effectiveness of the system is verified by conducting a performance comparison experiment between the robot and a professional musician. Results are briefly discussed and future works in the same direction are considered.

This article deals with a new robot for rehabilitation, DARTAGNAN, able to work in active or passive modes, on upper and lower human limbs. The presented robot has an hybrid serial/parallel structure with 6 degree of freedoms and a self-balanced mechanical structure. Although it is not an exoskeleton, DARTAGNAN behaves as such thanks to the customized software and to specific end-effectors, which can be appropriately connected to the forearm or lower leg. Thanks to the force/torque sensor the robot is able to treat the patient limb allowing him/her to (i) feel no effort in passive following mode, (ii) feel a specific effort in passive resistive mode or (iii) to actuate directly the limb in case of force deficit or spasticity for the active mode, while performing the exercise. Moreover, one of features of the robot is the possibility to directly calculate the limbs anthropometric parameters without the need of manual measurements to be set up before each treatment.

Exploiting underactuation and compliance in design and control of robotic hands represents a lively research topic. Reducing the actuators and increasing uncontrolled degrees of freedom, however may affect dexterity and control capabilities. Observing how humans use their hands in everyday grasping and manipulation tasks, it is evident that our “planner” is not always very precise, and we tend to use the environment to guide our finger trajectories to accomplish with the task. An interesting research challenge consists in providing also to robotics hands the ability of efficiently exploit environmental constraints. As a first step, it is necessary to extend to an augmented grasp system, including the hand, the object and the environment interaction, the main theoretical results and evaluation measures available for robotic grasping. In this paper we summarize an extended robotic grasp model including the interaction with the environment, and we apply to this system the evaluation of grasp stiffness. Some grasp evaluations for a simplified planar gripper are presented.

This paper formally documents a method that enables the scaling of displacements, elements of the group of rigid body motions, in all their six degrees of freedom individually in linear manner. The method incorporates in particular a suitable modulation method for the rotational part of a displacement. This simple, yet non-trivial function has not been reported in the literature as of today. The rotation scaling relies on the representation of a rotation in terms of its angle and its axis as given via Euler’s rotation theorem. It is argued that the overall workspace scaling falls in the class of ‘regularized linear functions’ in the tangent space.

One of the urgent problems facing the developers of multi-legged walking machines and robots is to increase the energy efficiency of the propulsion devices. During the transfer phase, the propulsion device detaches from the ground, rises to a certain height, and then descends to the ground. Having information on the nature of the profile of the bearing surface, it is possible to significantly reduce the distance traveled by the support and reduce the power required from the drive motor for carrying it according to the required law. The problem is set and a method is proposed for selecting the optimal energy losses for the motion of the walking robot for walking robot “Ortonog”, which was created in Volgograd at 2009–2011, by the criterion of minimum heat loss of energy when moving along an uneven surface. It is established that the mass-geometric parameters of the robot, characteristic of its drives and the geometric parameters of the supporting surface and the motion parameters influence the energy efficiency of the motion regime.

This study presents subject specific analysis at complementary time, frequency and phase plane domains of entire lower limb joint angular kinematic series during normal and modified gait. Healthy adult subject case study was assessed at human movement lab during normal gait, stiff knee gait and slow running. In vivo and non-invasive assessment was performed with eight camera Qualisys system at 100 Hz acquisition of cartesian coordinates of skin reflective markers at lower limb anatomical selected point. Inverse kinematics was performed using AnyGait v.0.92 and TLEM model to match the size and joint morphology of the stick-figure. Entire series of subject specific hip, knee and ankle joint angular displacements (θ), angular velocities (ω) and angular accelerations (α) at the sagittal plane were assessed at the time, frequency and phase plane domains providing useful information for development of subject specific rehabilitation equipment towards gait restoring. The main novelty and contribution of this study when compared with available literature consists on complementary approach of continuous entire signal analysis, time, frequency and phase analysis which can contribute for prescription of requisites for design of rehabilitation equipment towards normal gait of subjects with stiff knee gait and slow running.

This paper proposes a LQR controller for a double steering fast off-road mobile robot. The benefits of double steering robots are already well-known for low speed motion as the vehicle can track paths with high and small curvature. We extend this to dynamic motion (high speed) including wheel-ground slippage phenomena. We show that this double steering controller can ensure a good accuracy of path following at high and low speeds. This path tracking controller is evaluated by both numerical simulations and experimental tests in real conditions.

The expected adoption of robots in our society brings new technological challenges related to the utility and usability that these robots can provide to their potential users. In this paper, we discuss the main issues of current robot programming frameworks and interfaces for the development of usable and flexible end-user applications. In order to tackle these issues, we present Node Primitives (NEP), a robot programming framework aimed at enabling the creation of usable, flexible and cross-platform end-user programming (EUP) interfaces for robots. The applicability of NEP has been tested with the development of a block-based EUP interface.

This paper presents a new type of flying robot dedicated to grasping and manipulation of large size objects. The system can be basically presented as an aerial hand with four fingers actuated by four quadrotors whose arrangement permits the manipulation of the grasped object in the space. Each finger has two phalanges and is underactuated in order to adapt itself to the object size and shape. The opening/closing motion of each finger is actuated by the yaw motion of each quadrotor and transmitted through a non-backdrivable mechanism in order to enable the system to produce form-closed grasps. This stability criteria yields to secured grasps which do not rely on the capability of actuators nor on the contact friction between the phalanges and the object, and furthermore do not require any additional energy for gripping during the flight. The present paper gives guidelines to optimize the geometric parameters of a planar aerial robot in order to maximize the robot’s manipulability and its capability to produce form-closed grasps.

This paper proposes a mathematical model for a 7-links anthropomorphic biped - composed with two identical legs, two feet and one trunk - with an orthosis attached to both thigh and calf, during walking. We consider a cyclic walking gait in the sagittal plane. In the first part, we obtain trajectories of the motion for different walking speeds. It confirmed the fact that much more torque is recquired in the stance leg than in the swing leg. The second part analyses the impact of an orthosis, which fully assists the knee over the walking. Simulations point out the fact that valid joints have to compensate the weight of the orthotic system and show the interest to assist the knee of the stance leg.

With the emergence of dual-arm industrial robots lately, many robotic companies view dual-arm robots as main stream of industrial robots in the future and have begun research and development on them. The main objective of this work is to study dual-arm robotic manipulation issues on assembly tasks, involving the manipulation methods of dual-arm robots and how perception (integration of various kinds of sensors) and the feature of compliance (integration of hardware mechanism and online identification) are used to implement assembly tasks.

This paper analyses the dynamics of an antagonistically actuated tensegrity mechanism. The mechanism is subject to gravity effects, which produce both stable and unstable equilibrium configurations. The workspace is shown to be not necessarily connected and its size depends on both the geometric, spring and actuator parameters of the mechanism. The antagonistic actuation forces, which are bounded, enable controlling both the stiffness and the position within certain limits. A computed torque control law is applied and simulations show interesting behaviors of the mechanism when the desired motion makes the mechanism jump between two connected components of the workspace.

Problems with uniqueness and high parametric sensitivity of solution of equations of motion, encountered in the static friction regime, are addressed. Friction in joints of a multiple degree of freedom closed-loop kinematic chains is discussed. Two different models of friction are studied: the discontinuous Coulomb model with stiction regime represented in terms of additional constraints and the approximate Coulomb model, smoothed in the vicinity of zero relative velocity. Origins of non-uniqueness and high sensitivity are investigated; the questionable credibility of the stiction regime simulation results is pointed out. Two examples are provided to illustrate the discussed issues.

This paper describes a generalization of the virtual redundant axis (VRA) method previously introduced by the authors for robust, position-controlled inverse kinematics of serial robots passing through or close to a singular configuration. While the VRA method regards only one body- or inertially-fixed virtual axis, the new approach can handle multiple “floating” virtual axes which are smooth functions of the robot configuration and are not necessarily fixed to the robot links or base. This allows for robustly negotiating inverse kinematics even in the presence of multiple singularities with a single parameter set, while concentrating tracking errors solely in the locked directions. The paper discusses the theory, including a novel Jacobian form with non body-fixed joint axes, validating its performance on a real industrial robot Kuka KR15/2 and comparing it to the weighted damped least squares method (WDLS).

The synthesis of hand exoskeletons for rehabilitation is a challenging theoretical and technical task. A huge number of solutions have been proposed in the literature. Most of them are based on the concept to consider the phalanges of the finger as fixed to some links of the exoskeleton mechanism. This approach makes the exoskeleton synthesis a difficult problem that compels the designer to devise approximate technical solutions which, frequently, reduce the efficiency of the rehabilitation system and are rather bulky.This paper proposes a different approach. Namely, the phalanges are not fixed to some links of the exoskeleton, but they can have a relative motion, with one or two degrees of freedom when planar systems are considered. An example is presented to show the potentiality of this approach, which makes it possible: (i) to design glove-like exoskeletons that only approximate the human finger motion; (ii) to leave the fingers have their natural motion; (iii) to adapt a wider range of patient hand sizes to a given hand exoskeleton.

Synchronization tasks of robotic manipulators with moving objects are not only required to be solved in real-time but also in an optimal fashion. This paper considers optimal trajectory planning problems that are parameterized by the final state of the manipulator. For a sudden change of the desired final state, the trajectory needs to be replanned in real-time. Sensitivities of the optimal solution to a nominal problem w.r.t. the final state parameters are utilized to compute a nearly optimal real-time approximation of the solution to problem with perturbed parameters. Admissibility of the solution, i.e. satisfaction of constraints, is ensured by an iterative method supporting a variable active set of constraints. The efficacy of the proposed method is demonstrated in simulation and experiment.

This paper presents a method for the human ankle joint kinematic measurement by using a mechanical linkage having 6 d.o.f., all equipped with position sensors. The device allows a complete identification of all kinematic parameters of the joint which can be expressed here in the form of the Instantaneous Helical Axis and gait portraits for 3 principal rotations of the joint: dorsi/planta flexion, pronation/supination and internal/external rotation. A symmetry study of the 2 ankles during walking was also realized which highlighted the asymmetrical nature of the human gait as the FFT analysis gives a significant difference in amplitude between the rotational velocity vectors of the 2 ankle joints.