This paper presents a method for the automatic generation of robotic assembly sequences based on the verification of a disassemblability for a part to be disassembled. To derive the disassemblability of a candidate part, we first inference collision free assembly directions by extracting separable directions for the part and calculating the separability which gives information as to how the part can be easily separated. Using the result, we determine the disassemblability defined in terms of the separability and stability costs. The separability cost represents a facility of the part disassembly operation, while the stability cost represents a degree of the stability for the base assembly motion which gives a guidance for design of jig/fixture. The proposed method provides a potentially local optimal solution for finding a cost effective assembly plan, and the feasibility of generating robot motion programs by evaluating separable directions in flexible manufacturing applications. A case study is given to illustrate the concepts and procedure of the proposed scheme.

The objective of this experimental research was to determine whether or not an integrated vision and robotic arm system, using grey level detection and a transportation model heuristic, could plan and execute its own point-to-point navigation. The heuristic is the shortest-route algorithm which deals with finding the smallest possible distance to visit multiple locations only once. The only point taught by a human operator was a starting position and all other points within the workspace, to be visited by the robotic arm to perform tasks such as screw-fastening, adhesive application and detailed inspection of assemblies, were learned by the system.

A method of planning crab gaits for a quadruped walking robot is proposed. In the conventional gait study, one of the major concerns has been the foothold selection based on a prescribed body motion, but, in the paper, we consider the body motion planning problem under the assumption that a set of irregular footholds is given, and propose a hierarchical strategy. The strategy can be divided into three stages: first, a feasible range of the body movement is sought under the kinematic limit and the stability constraint. Next, a swing-leg sequence is selected with the aid of a proposed measure of traversability. Finally, an optimal sequence of the body motion is planned by the proposed procedure of otpimizing the gait stability margin. To verify the efficiency of the proposed method, simulation results are presented.

The problem of incremental terrain acquisition is addressed in this paper. Through a systematic planning of movements in an unknown terrain filled with polygonal obstacles, a sensor-based robot is shown to be able to incrementally build the entire terrain model; the model will be described in terms of visibility graph and visibility window. The terrain model is built area by area without any overlapping between explored areas. As a consequence, the terrain is obtained as a tessellation of disjoint star polygons. And the adjacency relations between star polygons are represented by a star polygon adjacency graph (SPAG graph). The incremental exploration process consists of two basic tasks: local exploration and exploration merging. Useful lemmas are derived for these two tasks and, then, the algorithms for the tasks are given. Examples are used to illustrate the algorithms. Two strategies for planning robot movements in the unknown terrain environment are suggested and compared. They are the depth-first search and the breadth-first search applied to the SPAG graph. Finally, the performance evaluation of the method and comparison with some existing methods are presented.

Automated methods are developed to classify a robot's kinematic type and select an appropriate library inverse-kinematic solution based on this classification. These methods automatically generate DenavitHartenberg joint frame parameters, given any frame representation that can mathematically be represented as a homogeneous transformation.

To reduce the number of closed-form inverse-kinematics solutions required for a broad class of serial robots, additional methods account for differences in robot zero state, base frame location, and joint polarity. Further generalization results from using joint frame decoupling to map lower degree-of-freedom robots into the inverse-kinematics solutions of higher degree-offreedom robots.

The paper proposes a concept of a movement diagnostics system of enhanced performances. The existing system, using electromyographic (EMG) and ground reaction force signals as inputs, has been applied primarily to sports locomotion evaluation and testing, providing suitable quantitative criteria. The proposed enhancement relies on movement simulation facility, incorporating subject-specific anatomical and functional data on the neuro-musculo-skeletal system of extremities. Reflecting the principles of anthropomorphic robotics, the computer graphics simulation is conceptualized to generate artificial (synthesized) movement patterns in virtual reality.

By comparing synthesized and measured signals the suitable feedback information is provided that may be used to enhance the overall system's performance. Using the power of computer technology and computer graphics, a user-friendly system of high ergonomic potentials is thus devised, in order to enhance the quality and efficacy of diagnostic procedure. The bioengineering principles used and the merging of measurement with simulation in the process of operation define a powerful instrument applicable to both healthy and pathological movement assessment.

This paper presents an example of the application of geometric and algebraic approaches to the inverse kinematics problem of four-link robot manipulators. A special arm configuration of the robot manipulator is employed for solving the inverse kinematics problem by using the geometric approach. The obtained joint variables as angular positions are defined in the form of cubic polynomials. The other kinematic parameters of the joints, such as angular velocities and angular accelerations, are the time derivatives of these polynomials. It is evident that there is no definite difference between the results of the two approaches. Consequently, if an appropriate arm configuration for the geometric approach can be established, the inverse kinematics can be solved in a simpler and shorter way.

Methods are described for computer analysis of image-data from a coded-stripe rangefinder. The main objective is to find vertex coordinates and connectivity information for a polyhedral object, enabling it to be represented by a wire-frame model. For each of several rangefinder viewpoints, the data is processed to extract three-dimensional edge and vertex positions. The emphasis is on estimation techniques that make good use of fairly sparse data-points. Results from different viewpoints are merged to produce a 3D model of the object.

A new vision system architecture has been developed to support the visual navigation of an autonomous mobile robot. This robot is primarily intended for urban park inspection, so it should be able to move in a complex unstructured environment. The system consists of various modules each ensuring a specific task involved in autonomous navigation. Task coordination focuses on the central module called the supervisor which triggers each module at the time appropriate to the current situation of the robot. Most of the processing time is spent with the scene exploration module which is based on the Hough transform to extract the dominant straight features. This module operates in two modes: the initial phase which forms the type of processing applied to the first image acquired in order to initiate navigation, and the continuous following mode which ensures the processing of subsequent images taken at the end of the blind distance. In order to rely less on visual data, a detailed map of the environment has been established, and an algorithm is used to make a scene prediction based on robot position provided by the localization system. The predicted scene is used to validate the objects detected by the knowledge base. This knowledge base uses the acquired and predicted data to construct a scene model which is the main element of the vision system.