Mobile robots in field environment travel not only on even terrain but also on uneven or sloped terrain. Practical methods for preventing turnover of the mobile robot are essential since the turnover of the mobile robot is very perilous. This paper proposes an efficient algorithm for preventing turnover of a mobile robot on uneven terrain by controlling linear acceleration and rotational velocity of the mobile robot. The concept of the modified zero moment point (ZMP) is proposed for evaluating the potential turnover of the mobile robot. Also, the turnover stability indices for linear acceleration and rotational velocity are defined with the modified ZMP. The turnover stability space (TSS) with turnover stability indices is presented to control the mobile robot in order to avoid turnover effectively. Finally, the feasibility and effectiveness of the proposed algorithm are verified through simulations conducted on a three-wheeled mobile robot.

In this paper, the robust robot localization problem with respect to uncertainties on environment features is formulated in a stochastic setting, and an extended Kalman filtering approach is proposed for the integration of odometric, video camera, and sonar measures. The environment is supposed to be only partially known, and a probabilistic method for sensor data fusion aimed at increasing the environment knowledge is considered.

In this paper, a novel spherical parallel manipulator and its isotropic design is introduced. This manipulator has good accuracy and relatively a larger workspace which is free of singularities. Utilizing spherical configuration the forward position problem is solved by equivalent angle–axis representation and Bezout's method which leads to a polynomial of degree 8. Two examples are given, one for isotropic and one for nonisotrpoic design. The first case results in eight real solutions, therefore, the polynomial being minimal. Using invariant form, we study acceleration analysis, conditions for singularity and find infinite isotropic structures. Accuracy and workspace analysis are also performed and are shown to have good global conditioning index and relatively large workspace. Using isotropic design and singularity requirements, we show the workspace of isotropic design is free of singularity.

Flexible manipulators have received wide attention because they surpass their rigid counterparts in many criteria. Unfortunately, traditional controller design methods for flexible arms implemented by human experts are usually tedious and intractable. In order to improve system behaviour, this paper proposes schemes based on genetic algorithms (GAs) which optimise the parameters of a fuzzy logic controller for a robotic manipulator with two-link flexibility and two-joint elasticity. Two alternative GA-optimised schemes are simulated and their control behaviour is compared with that of human-expert-designed controller.

This paper provides a comprehensive description of a new method of factorization for the Coriolis/centripetal matrix. In the past three decades, studies on dynamics have rapidly developed through the efforts of many researchers in the field of mechanics. While direct methods for deriving the Coriolis/centripetal matrix are well known and have been widely used in the last century, the entries of this matrix were always obtained by means of the Christoffel symbols of first kind. Startling techniques for deriving dynamic equations of robot manipulators first appeared about 30 years ago. Since then, much has been done to refine and develop the method, but it is still a highly active field of research, with many outstanding problems, both theoretical and in applications. This work presents, in a unitary frame and from a new perspective, the main concepts and results of one of the most fascinating aspects of mechanics, namely the factorization of structures, and offers the reader another point of view concerning a possible way to approach the Coriolis/centripetal matrix. It aims to study a theory of representation for such a matrix based on an elegant method of fundamental matrices. The paper is intended to be self-contained by presenting complete properties emerging from these novel structures. This work is useful not only to researchers in mechanics, but also to control engineers who are interested in learning some of the mechanical modeling. Toward this end, the paper provides numerical examples, as well as practical adaptive applications for modern designers to use at the system level.

This paper presents a passive dynamic walking model with toed feet that can walk down a gentle slope under the action of gravity alone. The model is the simplest of its kind with a point mass at the hip and two rigid legs each hinged at the hip on the one end and equipped with toed foot on the other end. We investigate two cases of the model, one with massless legs and another with infinitesimal leg masses. Rotation of the stance foot about the toe joint is initiated by ankle-strike, which is caused by the inelastic collision of the stance leg with a stop mounted on the stance foot. Numerical simulations of walking show that larger step lengths, higher speeds, stability, and energy efficiency can be achieved than what is achievable by a point-feet walker of same hip mass and leg lengths. Period-two gait of a point-feet walker is compared with period-one gait of the toed-feet walker and the mechanism responsible for achieving longer step lengths is described. It is shown that the advantage of the proposed walker comes from its relation to arc-feet walker. The characteristics of deterministic gait with infinitesimal leg masses is compared with that of nondeterministic gait with zero leg masses. It is shown that deterministic gait does not give maximum speed and efficiency compared to nondeterministic gait with swing leg control. Finally, active dynamic walking of the proposed walker is discussed.

People detection and tracking are essential capabilities in order to achieve a natural human–robot interaction. A great portion of the research in that area has been focused on monocular techniques. However, the use of stereo vision for these purposes concentrates a great interest nowadays. This paper presents a multi-agent system that implements a basic set of perceptual-motor skills providing mobile robots with primitive interaction capabilities. The skills designed use stereo and ultrasound information to enable mobile robots to (i) detect an interested user who desires to interact with the robot, (ii) keep track of the user while they move in the environment without confusing them with other people, and (iii) follow the user along the environment avoiding obstacles in the way. The system presented has been evaluated in several real-life experiments achieving good results and real-time performance on modest computers.

This study concerns the design and prototype of four different mobile robot platforms for rescue robot operations after an earthquake. At first, a test field is constructed to represent a mildly damaged earthquake zone. The test field consists of eight different sections: sand, gravel, ditch, water, bridge, incline, decline, and turn. The mechanical structure, electronics, software, communication, and possible sensory systems are explained. After the robots are manufactured, they are physically tested for their performance in the test field for 18 different parameters. The test results show the effective body structure. Challenges of the rescue robot design are explained and future expectations are given.

The robotic motion planning criteria has evolved from kinematics to dynamics in recent years. Many research achievements have been made in dynamic motion planning, but the externally applied loads are usually limited to the gravity force. Due to the increasing demand for generic tasks, the motion should be generated for various functions such as pulling, pushing, twisting, and bending. In this paper, a comprehensive form of equations of motion, which includes the general external loads applied at any point of branched tree structures, is implemented. An optimization-based algorithm is then developed to generate load-effective motions of redundant tree-structured systems for generic tasks. A highly articulated dual-arm human model is used to generate different effective motions to sustain different external load magnitudes. The results also provide a new scientific insight of human motion.

This paper describes real-time computer vision algorithms for detection, identification, and tracking of moving targets in video streams generated by a moving airborne platform. Moving platforms cause instabilities in image acquisition due to factors such as disturbances and the ego-motion of the camera that distorts the actual motion of the moving targets. When the camera is mounted on a moving observer, the entire scene (background and targets) appears to be moving and the actual motion of the targets must be separated from the background motion. The motion of the airborne platform is modeled as affine transformation and its parameters are estimated using corresponding feature sets in consecutive images. After motion is compensated, the platform is considered as stationary and moving targets are detected accordingly. A number of tracking algorithms including particle filters, mean-shift, and connected component were implemented and compared. A cascaded boosted classifier with Haar wavelet feature extraction for moving target classification was developed and integrated with the recognition system that uses joint-feature spatial distribution. The integrated smart video surveillance system has been successfully tested using the Vivid Datasets provided by the Air Force Research Laboratory. The experimental results show that system can operate in real time and successfully detect, track, and identify multiple targets in the presence of partial occlusion.

To evaluate our three-axis tactile sensor developed in preceding papers, a tactile sensor is mounted on a robotic finger with 3-degrees of freedom. We develop a dual computer system that possesses two computers to enhance processing speed: one is for tactile information processing and the other controls the robotic finger; these computers are connected to a local area network. Three kinds of experiments are performed to evaluate the robotic finger's basic abilities required for dexterous hands. First, the robotic hand touches and scans flat specimens to evaluate their surface condition. Second, it detects objects with parallelepiped and cylindrical contours. Finally, it manipulates a parallelepiped object put on a table by sliding it. Since the present robotic hand performed the above three tasks, we conclude that it is applicable to the dexterous hand in subsequent studies.

A novel 4-dof 2SPS+SPR parallel kinematic machine is proposed, and its kinematics, statics, and workspace are studied systematically. First, the geometric constrained equations are established, and the inverse displacement kinematics is analyzed. Second, the poses of active/constrained forces are determined, and the formulae for solving inverse/forward velocities are derived. Third, the formulae for solving inverse/forward accelerations are derived. Finally, a workspace is constructed and its active/constrained forces are solved. The analytic results are verified by its simulation mechanism to be consistent with the calculated ones.

This paper presents a novel model of snake-like robots based on a spatial linkage mechanism. The reasonable structural parameters of the mechanism are obtained by performing a kinematic simulation. Then the kinematics of the spatial linkage mechanism is developed and the motor angles of the robot for performing lateral undulation are analyzed based on the Serpenoid curve. The torque of servomotors at each moment is also obtained. The experiments detailed in this paper confirm that the robot is of the ability to realize several motion modes, including lateral undulation, left and right turning motions, and uplifting of the head.

This work presents an analysis of the mechanics of biped running with a heuristic control strategy for stable running. A tractable model of the mechanics is presented through principles, measures, and a time-distributed perspective. This model motivates the control strategy and provides a basis for extending the strategy to more general systems. The control strategy consists of a simple set of four rules, where the key rule considers the leg length upon liftoff. The Steady-State Index is proposed as a characterization of biped running. It identifies and relates the parameters affecting running and has broad applicability to both biological and robotic systems.