Passive walking robots can walk on a slight downward slope powered only by gravity. We propose a novel control strategy based on forced entrainment to stabilize a three-dimensional quasi-passive walking robot in uphill and level walking by using torso control in the frontal plane and synchronization of lateral motion with swing leg motion. We investigated the robot's walking energy efficiency, energy transformation, and transfer in simulation. The results showed that the proposed method is effective and energy-efficient for uphill and level walking. The relationship between energy utilization rate of actuation and energy efficiency of the robot was revealed, and mechanical energy transformation and transfer were characterized.

A new robust neural sliding mode (RNSM) tracking control scheme using radial basis function (RBF) neural network (NN) is presented for MEMS z-axis gyroscope to achieve robustness and asymptotic tracking error convergence. An adaptive RBF NN controller is developed to approximate and compensate the large uncertain system dynamics, and a robust compensator is designed to eliminate the impact of NN modeling error and external disturbances for guaranteeing the asymptotic stability property. Moreover, another RBF NN is employed to learn the upper bound of NN modeling error and external disturbances, so the prior knowledge of the upper bound of system uncertainties is not required. All the adaptive laws in the RNSM control system are derived in the same Lyapunov framework, which can guarantee the stability of the closed loop system. Comparative numerical simulations for an MEMS gyroscope are investigated to verify the effectiveness of the proposed RNSM tracking control scheme.

Human–robot interaction is an emerging area of research where a robot may need to be working in human-populated environments. Human trajectories are generally not random and can belong to gross patterns. Knowledge about these patterns can be learned through observation. In this paper, we address the problem of a robot's social awareness by learning human motion patterns and integrating them in path planning. The gross motion patterns are learned using a novel Sampled Hidden Markov Model, which allows the integration of partial observations in dynamic model building. This model is used in the modified A* path planning algorithm to achieve socially aware trajectories. Novelty of the proposed method is that it can be used on a mobile robot for simultaneous online learning and path planning. The experiments carried out in an office environment show that the paths can be planned seamlessly, avoiding personal spaces of occupants.

Tethered mobile robots are ideal for electrically noisy environments and for time-consuming tasks that require robust data communication and uninterrupted power delivery. However, tethers may become entangled in cluttered environments, leading to immobilisation and consequent mission failure. This work addresses real-time monitoring of tethers to detect tether entanglement, perform disentanglement through tether following and localise within line of sight. Experimental hardware is proposed to implement the tether monitoring techniques. Experiments are performed for single and dual mobile robots to search a target environment and entanglement detection is shown to be successful using quantitative metrics such as mean localization error.

This paper discusses about a proposed solution to the obstacle avoidance problem in multi-robot systems when applied to active target tracking. It is explained how a nonlinear model predictive formation control (NMPFC) previously proposed solves this problem of fixed and moving obstacle avoidance. First, an approach is presented which uses potential functions as terms of the NMPFC. These terms penalize the proximity with mates and obstacles. A strategy to avoid singularity problems with the potential functions using a modified A* path planning algorithm was then introduced. Results with simulations and experiments with real robots are presented and discussed demonstrating the efficiency of the proposed approach.

This paper proposes a new dynamic planning navigation strategy for use with mobile terrestrial robots. The strategy was applied to situations in which the environment and obstacles were unknown. After each displacement event, the robot replanned its route using a control algorithm that minimized the distance to the target and maximized the distance between the obstacles. Using a spatial localization sensor and a set of distance sensors, the proposed navigation strategy was able to dynamically plan optimum routes that were free of collisions. Simulations performed using different types of environment demonstrated that the technique offers a high degree of flexibility and robustness, and validated its potential use in real applications involving mobile terrestrial robots.

The weightless planar 2R underactuated manipulators with passive last joint are considered in this paper for investigating a feasible method to stabilize the system, which is a second-order nonholonomic-constraint mechanical system with drifts. The characteristics including the controllability of the linear approximation model, the minimum phase property, the Small Time Local Controllability (STLC), the differential flatness, and the exactly nilpotentizable properties, are analyzed. Unfortunately, these negative characteristics indicate that the simplest underactuated mechanical system is difficult to design a stable closed-loop control system. In this paper, nilpotent approximation and iterative steering methods are utilized to solve the problem. A globally effective nilpotent approximation model is developed and the parameterized polynomial input is adopted to stabilize the system to its non-singularity equilibrium configuration. In accordance with this scheme, it is shown that designing a stable closed-loop control system for the underactuated mechanical system can be ascribed to solving a set of nonlinear algebraic equations. If the nonlinear algebraic equations are solvable, then the controller is asymptotically stable. Some numerical simulations demonstrate the effectiveness of the presented approach.

Elbows and branches in pipelines produce unique image patterns, which can be used as landmarks for autonomous navigation inside the pipelines. This paper presents two landmark detection methods, known as shadow-based method and laser projection method. The first method uses the landmark's unique patterns of shadow produced by the robot's illuminator. The other method exploits special line features generated by its own line-laser beam projector. The basic algorithms for extracting the landmarks are given and special sensor mechanisms are addressed respectively. Finally, the detection performances of each method are validated in various pipeline conditions by using an in-pipe robot, called MRINSPECT-V.

Fault tolerance is a very important issue for legged robots, especially in some harsh environments. One of the most fragile parts is the actuation system. There are two common faults of robot actuators: (1) the motor is locked and could not move anymore; (2) the motor is uncontrollable and can be treated as a passive joint. In this paper, we first discuss all fault combinations of a single leg of a hexapod walking robot with parallel-parallel mechanism topology. Then, the leg tolerable criterion is brought out, which defines whether a leg is fault tolerant. After that, the fault tolerance of the whole robot is researched, and we found that the robot can walk with one tolerable leg or two opposite tolerable legs. Finally, relative simulation results are given, which show the robot walk with one or two broken legs.

A method for motion/force control of robot arms with model uncertainties is presented. Tracking control of complex trajectories is guaranteed using a Lyapunov approach with high-precision performance ensured using a particle swarm optimization (PSO) algorithm. Tracking performance and robustness are simulated for a robotic device for limb rehabilitation that is designed to be adapted easily to different subjects by considering model parameter uncertainties. Controller parameters are optimized offline using the PSO algorithm with Lyapunov stability conditions considered as inequality constraints. Using the control scheme, the robot can guide limbs on smooth and non-smooth trajectories, under model uncertainties and measurement noise.

The goal of this research is to develop a low-cost face robot which has a lower degree-of-freedom facial expression mechanism. Many designs of facial robots have been announced and published in the past. Face robots can be classified into two major types based on their respective degrees of freedom. The first type has various facial expressions with higher degrees of freedom, and the second has finite facial expressions with fewer degrees of freedom. Due to the high cost of the higher-degree-of-freedom face robot, most commercial face robot products are designed in the lower-degrees-of-freedom form with finite facial expressions. Therefore, a face robot with a simplified facial expression mechanism is proposed in this research. The main purpose of this research is to develop a device with a lower degree-of-freedom mechanism that is able to generate many facial expressions while keeping one basic mouth shape variation. Our research provides a new face robot example and development direction to reduce costs and conserve energy.

This paper introduces the design and the optimization of a probe holder robot for tele-echography applications. To define its kinematic architecture, an approach based on motion capture of an expert's gestures during ultrasound examinations was proposed. The medical gestures analyzed consisted of ultrasound probe movements and were used to characterize the kinematic specifications of the proposed manipulator. The selected architecture was a Spherical Parallel Mechanism (SPM) with 3 degrees of freedom (DoF) and its optimal synthesis was performed using real-coded Genetic Algorithms (GA). The optimization criteria and constraints were established thanks to the collaboration of medical experts and were successively formulated and solved using mono-objective and multi-objective functions.

The inverse dynamics model of a novel (3-UPU)+(3-UPS+S) serial–parallel manipulator (S-PM) formed by a 3-UPU PM and a 3-UPS+S PM connected in serial is studied in this paper. First, the inverse position, velocity, and acceleration of this S-PM are studied systematically. Second, the velocity mapping relations between each component and the terminal platform of (3-UPU)+(3-UPS+S) S-PM are derived. Third, the dynamics model of the whole (3-UPU)+(3-UPS+S) S-PM is established by means of the principle of virtual work. The process for establishing the dynamics model of this S-PM is fit for other S-PMs.

The most important applications of many computer vision systems are based on robust features extraction, matching and tracking. Due to their extraction techniques, image features locations accuracy is heavily dependent on the variation in intensity within their neighbourhoods, from which their uncertainties are estimated. In the present work, a robust L∞ optimisation solution for monocular motion estimation systems has been presented. The uncertainty estimation techniques based on SIFT derivative approach and its propagation through the eight-point algorithm, singular value decomposition SVD and the triangulation algorithm have proved an improvement to the global motion estimation. Using monocular systems makes the motion estimation challenging due to the absolute scale ambiguity caused by projective effects. For this, we propose robust tools to estimate both the trajectory of a moving object and the unknown absolute scale ratio between consecutive image pairs. Experimental evaluations showed that robust convex optimisation with the L∞ norm under uncertain data and the Robust Least Squares clearly outperform classical methods based on Least Squares and Levenberg-Marquardt algorithms.