A potential panel method is proposed to solve the collision-free pathplanning problem for a free-flying robot operating in an obstacle filled 3-Denvironment. The problem is solved in three steps: (1) transform the arbitraryshaped obstacles in the 3-D workspace into simple convex polyhedrons; (2)generate an artificial potential field using the 3-D panel method in the 3-Dworkspace; (3) generate a streamline from the starting position towards the goalposition in the artificial potential field. The computational complexity of thepertinent algorithms justify the efficiency of the approach and itsapplicability in real-time. Simulation results illustrate the potential of theproposed approach. The reported research is the outgrowth of the 2-D method,already published.

A link position tracking controller is formulated for an n-link, rigid,revolute, serially-connected robot. The controller generates torque commands tothe individual robot links based on adaptive estimates of the system parametersand measurements of only link positions. A filtering technique, based on thelink position signal, is used to alleviate the need for velocity measurements. Acomplete development of the controller is presented along with a proof ofsemiglobal asymptotic link position-velocity tracking performance. Experimentalvalidation of the proposed controller on the Integrated Motion Inc. (IMI)two-link direct drive robot is also presented. Several extensions to the basiccontroller are described that consider the use of fixed parameterestimates.

We constructed a high reliability parts-picking device with an active and multi-sensor visual system, that picks up a part from randomly stacked parts. The system determines the orders in which to pick up the parts. To determine the order, we introduce concepts of risks of collapsing a stack by picking, and the ease of manipulator operations. The system examines if the risks exist and estimates the ease. We show that the parts-picking system can achieve high reliability by determining the orders, using the effects of the examination and the estimation. We also give some experimental results.

In this paper, we propose a two-loop structure to transform and stabilize the kinematic model of a nonholonomic mobile robot with two independently driven wheels. This two-loop structure consists of input-output pseudolinearization and gain scheduling techniques. A comparison with previous methods is included. The main contribution of this paper is to apply a input-output pseudolinearization transformation method and to use an effective pole-assignment strategy for stabilizing a mobile robot with two independently driven wheels. The proposed method has demonstrated superiority over previous methods.

A graphical robot programming system has been developed. This system with a graphical interface is user-friendly and easy-to-learn for low-skill users. It has been developed as a prototype system under a project by the Japan Robot Association (JARA) since 1994. The system runs on a personal computer and consists of a graphical user interface and an editing system. It is designed for programming an arc welding robot in small batch production and is expected to provide low-skill users with a means to use industrial robots with ease.

A task based approach to the issue of redundant robots starts from thepremise that there are obstacles that cannot be removed from the working areaand which therefore must be avoided. This statement produces the requirement fora device with a certain degree of mobility, and stresses the need to ensure thatthe aim is twofold: reach the goal and avoid obstacles. But avoiding obstaclesis not the same objective as keeping as far away from an obstacle as possible;the primary goal is still to reach the target. In fact humans use soft contactto reach targets that are at the periphery of their reach. This soft distributedcontact has the effect of smoothing the surface of the object and hence there isan element of only being interested in obstacle detail at the appropriate scaleto achieve the task.

This paper describes a new approach to collisionavoidance based on using a global path finding algorithm, in this case usingLaplacian potential fields, in conjunction with a simple local geometricallybased algorithm for avoiding obstacles and maximising the use of manoeuvringspace in a manner which is not limited by digital computation resolution issues.This extra technique is in some ways analogous to the human soft contactapproach.

Three examples are presented to illustrate the robustness ofthe algorithm. In order to be able to compare results with other techniques, anenvironment measurement scheme is defined which gives an indication of thedifficulty of the trajectory being followed.

Syntax-directed development is a software development method in which, the syntax of the input of the application plays a central role. The syntax description forms a frame on which semantic actions, attributes, local and global information can be attached. Much research has been done to develop theories and algorithms needed for robots to process information and interact with the environment. This paper describes how LL(1) descriptions can be used to produce automatically motion control software, object recognizers and human-computer interfaces.

Several descriptions of redundancy are presented in the literature, often from widely different perspectives. Therefore, a discussion of these various definitions and the salient points would be appropriate. In particular, any definition and redundancy needs to cover the following issues; the difference between multiple solutions and an infinite number of solutions; degenerate solutions to inverse kinematics; task redundancy; and the distinction between non-redundant, redundant and highly redundant manipulators.

The UK Magnetics Society's one-day seminar Advances inInstrumentation for Magnetics Measurements was held at the University ofWales Cardiff on 10 July 1996. Ten presentations were made during the eventwhich was attended by over 65 participants.

A robot has to be safe and reliable. An unreliable robot may become the cause of unsafe conditions, high maintenance costs, inconvenience, etc.

Over the years, in general safety and reliability areas various assessment methods have been developed, e.g. failure mode and effects analysis, fault tree analysis, and Markovian analysis. In view of these, this paper presents an overview of the most suitable robot safety and reliability assessment techniques.

The ORCCAD programming environment for robotic systems gathers control laws in continuous time at the low levels and discrete time logical aspects at higher levels. Based upon a formal definition of robotic actions, complex applications can be designed, verified and generated incrementally. The approach and tools prototypes have been validated through several applications.

In this paper, we propose an approach which ensures the dynamicstability of a biped robot called “BIPMAN”. It is based on the correction of thetrunk center of mass acceleration and on the distribution of the forces exertedby the limbs on the trunk. This latter is performed by means of a linearprogramming method (the simplex method). The retained criterion allows tooptimize force distribution as well as trunk roll and pitch angles. Weightingfactors are introduced into this criterion in order to define criteria adaptedto specific tasks. Modifying these factors is a solution to the task transitionproblem. Many simulation results are presented to demonstrate criteria andconstraints influences on dynamic stability. They lead us to introduce a newapproach called RTCA (Real Time Criteria andConstraints Adaptation). This relies on the analysis ofposition, velocity and acceleration vectors for criterion real time adaptationand on forces analysis for constraints real time adaptation. The RTCA approach is finally validated through simulations results for a specifictask.

Sensor based robotic systems are an important emerging technology. Whenrobots are working in unknown or partially known environments, they need rangesensors that will measure the Cartesian coordinates of surfaces of objects intheir environment. Like any sensor, range sensors must be calibrated. The rangesensors can be calibrated by comparing a measured surface shape to a knownsurface shape. The most simple surface is a plane and many physical objects haveplanar surfaces. Thus, an important problem in the calibration of range sensorsis to find the best (least squares) fit of a plane to a set of 3D points.

We have formulated a constrained optimization problem to determine the leastsquares fit of a hyperplane to uncertain data. The first order necessaryconditions require the solution of an eigenvalue problem. We have shown that thesolution satisfies the second order conditions (the Hessian matrix is positivedefinite). Thus, our solution satisfies the sufficient conditions for a localminimum. We have performed numerical experiments that demonstrate that oursolution is superior to alternative methods.