This paper presents a module for pedestrian detection from a moving vehicle in low-light conditions. The algorithm make use of a single far infrared camera based on a microbolometer. Images of the area ahead of the vehicle are analyzed to determine if any pedestrian might be in its trajectory. Detection is achieved by searching for distributions of temperatures in the scene similar to that of the human body. Those areas with an appropriate temperature, size, and position in the image are classified, by means of a correlation between them and some probabilistic models, which represents the average temperature of the different parts of the human body. Finally, those pedestrians found are tracked in a subsequent step, using an unscented Kalman filter. This final stage of the algorithm enables the algorithm to predict the trajectory of the pedestrian, in a way that does not depend on the movement of the camera. The aim of this system is to warn the vehicle's driver and reduce the reaction time in case an emergency break is necessary.

This work is concerned with trajectory tracking of robots when the electrical dynamics of the brushless DC motor actuators is considered. It is shown that proportional-derivative (PD) control with feedforward compensation, plus some additional terms to cope with the electrical dynamics, ensures state boundedness. Furthermore, tracking error converges to zero from arbitrarily large initial conditions if controller gains are correctly chosen. Under mild assumptions, this controller reduces to the well-known PD control with feedforward compensation when implemented according to torque control, a successful industrial practice. Thus, it is explained, for the first time, why this strategy works well in applications.

The formation of autonomous mobile robots to an arbitrary geometric pattern in a distributed fashion is a fundamental problem in formation control. This paper presents a new asynchronous, memoryless (oblivious) algorithm to the formation problem via distributed optimization techniques. The optimization minimizes an appropriately defined difference function between the current robot distribution and the target geometric pattern. The optimization processes are performed independently by individual robots in their local coordinate systems. A movement strategy derived from the results of the distributed optimizations guarantees that every movement makes the current robot configuration approaches the target geometric pattern until the final pattern is reached.

To enhance tactile display capability, we performed a series of experiments based on psychophysics in which we adopted a circle and eight kinds of ovals as stimuli of the convex line (hereafter tactile image). The circle's diameter is 10 mm, the minor axes of the eight ovals are decreased in 3% steps from the circles, and each major axis is adjusted so that the area of each oval equals the circle area. Human subjects judged which stimuli (presented on the left or the right tactile images) equal the circle. The experiment was conducted on both the finger pads and the thenar parts to compare their experimental sensitivity ratios with the sensitivity ratio estimated on the basis of mechanoreceptor density. The discrimination sensitivity for the thenar part is inferior to that for the finger pad in terms of the low density of the mechanoreceptor in the thenar part. The tactile image was mainly recognized by a slowly adaptive type I unit, but the first adaptive type II unit was not affected by specimen cooling due to discrimination precision. Subsequently, discrimination sensitivity was not different between the free and restricted haptic motions. Since this result implies that the tactile image is modified by a motor signal for the haptic motion to prevent blurring of the tactile image, perhaps the motion movement in the tactile display is changed based on the design convenience.

This paper investigates the kinematic calibration of a 3-DOF parallel mechanism based on the minimal linear combinations of error parameters. The error mapping function between the geometric errors and the output errors is formulated and the identification matrix is generated and simplified. In order to identify the combinations of error parameters, four theorems to analyze the columns of the simplified identification matrix are introduced. Then, an anti-disturbance index is presented to evaluate the identification performance. On the basis of this index, measurement strategy is developed and optimal measuring configurations are given. After external calibration, linear interpolation compensation is applied to improve the terminal accuracy further. Results of experiment show that the method used in this paper is effective and efficient, and the errors are convergent within two iterations generally. This method can be extended to other parallel mechanisms with weakly nonlinear kinematics.

In this paper, an experimental verification of antagonistic stiffness planning is presented for a 2-DOF parallel mechanism with four actuators. With 2-DOF force redundancy, the magnitude and direction of the stiffness enhancement can be controlled by the additional actuators, where the internal torques of the mechanism exist on the two-dimensional null space. In the experiments, the passive and active stiffness are measured, respectively, during endowing the external force at the end-effector. Two stiffness assignment methods for a given pathway are suggested and are verified by the experiments.

Mobile robot systems are being used more often in tasks that protect human operators from dangerous environments, but these benefits can be easily lost if the robots spend much of their time being repaired. This implies that any increment in their reliability will also improve their benefits. One way to achieve this is by adding redundant elements to the robot, but this adds complexity and cost to the design. On the other hand, cooperative mobile robots formed by members with the same basic structure provide a natural redundancy of elements, which may be used for reliability improvement. This work presents an architecture that takes advantage of the analytical and sensor redundancy present in groups of cooperative mobile robots in order to increase the reliability of the whole system. First, the design of the architecture is portrayed and the faults to be detected are described. The different layers of the system are then explained and analyzed using several simulations to test their capabilities and limitations. Finally, the experimental results on a group of small mobile robots are shown, validating the results delivered by simulations. These results show that the proposed architecture is able to detect and isolate correctly most of the faults tested.

Based on the terminal constraints system (TCS) and reciprocal screw theory, a novel method is presented to determine the isotropic configurations of limited degree-of-freedom (DOF) parallel manipulators. From the available physical meaning of isotropy, the criteria to determine the isotropic configurations can be transformed to investigate whether the TCS acting on the moving platform works equally well in all directions. From the TCS study, the simplest form of constraints system matrix can be obtained. Then the constraint condition number is defined to measure the isotropy of spatial parallel manipulator based on the TCS. This method not only avoids solving the Jacobian matrix for some complex structural parallel manipulators but also points out the physical meaning of isotropy, which indicates that the TCS acting on the moving platform works equally well in all directions. Three examples are employed to illustrate this method.

In this work a novel spatial hyper-redundant manipulator inspired in the motions of the worms is introduced. The displacement analysis is presented in a semi-closed form solution, whereas the velocity and acceleration analyses are carried out by means of the theory of screws. Among typical applications of most hyper-redundant manipulators, interesting biomechanical applications such as the simulation of the motion of the spine are available for this new artificial worm.

The generalized Jacobian matrix was introduced for dealing with end-effector control in space robots. One of the applications of this Jacobian is to be used in Jacobian transpose control to generate joint torques given end-effector position error. It would be misleading, however, to consider the transpose of this Jacobian as a mapping from end-effector force/moment to controlled joint torques for underactuated systems or floating base robots. This paper explains why it does not represent the mapping and provides a simple example. Later, the correct mapping is provided using the dynamically consistent Jacobian inverse and then a method to compute the actuated-joint torques is explained given the desired end-effector force. Finally, the effect of using the generalized Jacobian in the Jacobian transpose control is analyzed.

In parametric excitation walking, up-and-down motion of the center of mass restores mechanical energy and sustainable gait is generated. Not only walking performance but also walking ability strongly depends on the reference trajectory of the center of mass. In this paper, we propose an optimization method for the reference trajectory, in which the reference trajectory is confined to the quartic spline curves and the parameters of spline curves are optimized by a local search method usually used in combinatorial optimization. We apply the proposed method to a kneed biped robot and find some remarkably interesting results by numerical simulations.

From the viewpoint of the system's mechanical energy, the passive inverted pendulum model (PIPM) is proposed for the generation of more energy-efficient biped gait pattern. The generated walking pattern, based on the PIPM, enables the fully actuated biped robots to closely mimic the behavior of stable passive walking, so that it can have good energy-efficiency, which is the inherent advantage of the passive system. Furthermore, the pattern generation method is extended to any desired terrain as well. As for SHR-1, the first-generation biped robot of Shanghai Jiao Tong University, gait synthesis is clarified in detail. Finally, the walking experiments are carried out on SHR-1, and the effectiveness of the proposed pattern generation method is confirmed.

In this paper, a generalized form of the antipodal method from multi-finger grasping is presented and implemented for investigating the workspace of a wide range of planar wire-actuated parallel manipulators. Manipulators with distinct wire attachment points on the base and mobile platform are considered, in the absence and presence of external force. The examined workspaces are verified with the corresponding workspaces developed using static force analysis. By applying an external force, modeled as a wire for the antipodal method, the characteristics of the manipulator could be altered by enlarging its workspace in the direction of the applied force.

Cyclic coordinate descent (CCD) inverse kinematics methods are traditionally derived only for manipulators with revolute and prismatic joints. We propose a new numerical CCD method for any differentiable type of joint and demonstrate its use for serial-chain manipulators with coupled joints. At the same time more general and simpler to derive, the method performs as well in experiments as the existing analytical CCD methods and is more robust with respect to parameter settings. Moreover, the numerical method can be applied to a wider range of cost functions.

In this paper, an offline algorithm generates gaits for robotic biped running motion. A minimization procedure on a novel objective function produces initial joint velocities for the flight phase given the initial joint angles. These computed initial velocities ensure a reasonable starting point for the constrained dynamic optimization procedure to try and formulate a gait that minimizes energy consumption. The simulated robot runs even on ground and up stairs. Results demonstrate the rigid robot solution has a intuitively satisfying human-like motion with an energy consumption that appears reasonable for a rigid robot.

This paper focuses on the use of passive dynamics to achieve efficient walking with simple mechanisms. A torso is added to a biped walker; and hip actuators, instead of ankle actuators, are used. A numerical approach is used to find the optimal control trajectories. A comparison between the cost functions of simple feedback control and optimal control is presented. Next, springs are added to the biped walking model at the hip joints to demonstrate the advantage of hip springs in terms of energy cost and ground conditions. The comparison between the torque costs with and without hip springs indicates that hip springs reduce the torque cost, particularly at a high walking speed.